Simulace.info - User contributions [en]

Customer segmentation techniques

2016-01-24T22:11:49Z

Xtomp36: /* Customer segmentation in SimProcess */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

[[File:segment.jpg]]

Figure 1 - graphical illustration on how usage of customer segmentation effects business revenue due to development of additional products, which is only one of many possible outcomes.

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Usage segmentation==

Another rather simple approach of customer segmentation. There are two methods for usage segmentation either the customers are divided based on their weight of use or by time and place of usage. With the first method it is obvious that customers who buy more are more important to the business that the other ones. In here for example a “Pareto analysis” could be used to identify the top 20 % of most valuable customers. This method is normally used in business-to-business markets. The second method divides customers based on time and place. At different times customers may want different products available, so this method takes that into consideration. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 2 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 3 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Simulator and simulation game=
This part of the article describes two cases of customer segmentation usage, where simulation is a factor.

==Market Segmentation Simulator==
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

==Marketing Management Simulation Game==
This simulation game has some aspects of customer segmentation, so I find it worth mentioning in this article. It is developed by Cesim SimBrand and targets the marketing decision making process. It has 8 customer segments, 2 market areas and six different products. The game covers many others areas of marketing, than just customer segmentation, like: product life cycle management, positioning, after sales services, pricing, sales forecasting, competitor analysis and profitability. Its expected outcome is to help the participant to get a better understanding of the whole process. <ref> Marketing Management Simulation Game [online]. [cit. 2016-01-24]. Available at: http://www.cesim.com/simulations/cesim-simbrand-marketing-management-simulation-game </ref>

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T22:11:38Z

Xtomp36: /* Clustering approach */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

[[File:segment.jpg]]

Figure 1 - graphical illustration on how usage of customer segmentation effects business revenue due to development of additional products, which is only one of many possible outcomes.

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Usage segmentation==

Another rather simple approach of customer segmentation. There are two methods for usage segmentation either the customers are divided based on their weight of use or by time and place of usage. With the first method it is obvious that customers who buy more are more important to the business that the other ones. In here for example a “Pareto analysis” could be used to identify the top 20 % of most valuable customers. This method is normally used in business-to-business markets. The second method divides customers based on time and place. At different times customers may want different products available, so this method takes that into consideration. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 2 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Simulator and simulation game=
This part of the article describes two cases of customer segmentation usage, where simulation is a factor.

==Market Segmentation Simulator==
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

==Marketing Management Simulation Game==
This simulation game has some aspects of customer segmentation, so I find it worth mentioning in this article. It is developed by Cesim SimBrand and targets the marketing decision making process. It has 8 customer segments, 2 market areas and six different products. The game covers many others areas of marketing, than just customer segmentation, like: product life cycle management, positioning, after sales services, pricing, sales forecasting, competitor analysis and profitability. Its expected outcome is to help the participant to get a better understanding of the whole process. <ref> Marketing Management Simulation Game [online]. [cit. 2016-01-24]. Available at: http://www.cesim.com/simulations/cesim-simbrand-marketing-management-simulation-game </ref>

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T22:11:27Z

Xtomp36: /* Possible costumer segmentation techniques */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

[[File:segment.jpg]]

Figure 1 - graphical illustration on how usage of customer segmentation effects business revenue due to development of additional products, which is only one of many possible outcomes.

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Usage segmentation==

Another rather simple approach of customer segmentation. There are two methods for usage segmentation either the customers are divided based on their weight of use or by time and place of usage. With the first method it is obvious that customers who buy more are more important to the business that the other ones. In here for example a “Pareto analysis” could be used to identify the top 20 % of most valuable customers. This method is normally used in business-to-business markets. The second method divides customers based on time and place. At different times customers may want different products available, so this method takes that into consideration. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Simulator and simulation game=
This part of the article describes two cases of customer segmentation usage, where simulation is a factor.

==Market Segmentation Simulator==
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

==Marketing Management Simulation Game==
This simulation game has some aspects of customer segmentation, so I find it worth mentioning in this article. It is developed by Cesim SimBrand and targets the marketing decision making process. It has 8 customer segments, 2 market areas and six different products. The game covers many others areas of marketing, than just customer segmentation, like: product life cycle management, positioning, after sales services, pricing, sales forecasting, competitor analysis and profitability. Its expected outcome is to help the participant to get a better understanding of the whole process. <ref> Marketing Management Simulation Game [online]. [cit. 2016-01-24]. Available at: http://www.cesim.com/simulations/cesim-simbrand-marketing-management-simulation-game </ref>

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T22:11:13Z

Xtomp36: /* A priory segmentation */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Usage segmentation==

Another rather simple approach of customer segmentation. There are two methods for usage segmentation either the customers are divided based on their weight of use or by time and place of usage. With the first method it is obvious that customers who buy more are more important to the business that the other ones. In here for example a “Pareto analysis” could be used to identify the top 20 % of most valuable customers. This method is normally used in business-to-business markets. The second method divides customers based on time and place. At different times customers may want different products available, so this method takes that into consideration. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Simulator and simulation game=
This part of the article describes two cases of customer segmentation usage, where simulation is a factor.

==Market Segmentation Simulator==
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

==Marketing Management Simulation Game==
This simulation game has some aspects of customer segmentation, so I find it worth mentioning in this article. It is developed by Cesim SimBrand and targets the marketing decision making process. It has 8 customer segments, 2 market areas and six different products. The game covers many others areas of marketing, than just customer segmentation, like: product life cycle management, positioning, after sales services, pricing, sales forecasting, competitor analysis and profitability. Its expected outcome is to help the participant to get a better understanding of the whole process. <ref> Marketing Management Simulation Game [online]. [cit. 2016-01-24]. Available at: http://www.cesim.com/simulations/cesim-simbrand-marketing-management-simulation-game </ref>

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T22:10:36Z

Xtomp36: /* A priory segmentation */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

[[File:segment.jpg]]

Figure 1 - graphical illustration on how usage of customer segmentation effects business revenue due to development of additional products, which is only one of many possible outcomes.

==Usage segmentation==

Another rather simple approach of customer segmentation. There are two methods for usage segmentation either the customers are divided based on their weight of use or by time and place of usage. With the first method it is obvious that customers who buy more are more important to the business that the other ones. In here for example a “Pareto analysis” could be used to identify the top 20 % of most valuable customers. This method is normally used in business-to-business markets. The second method divides customers based on time and place. At different times customers may want different products available, so this method takes that into consideration. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Simulator and simulation game=
This part of the article describes two cases of customer segmentation usage, where simulation is a factor.

==Market Segmentation Simulator==
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

==Marketing Management Simulation Game==
This simulation game has some aspects of customer segmentation, so I find it worth mentioning in this article. It is developed by Cesim SimBrand and targets the marketing decision making process. It has 8 customer segments, 2 market areas and six different products. The game covers many others areas of marketing, than just customer segmentation, like: product life cycle management, positioning, after sales services, pricing, sales forecasting, competitor analysis and profitability. Its expected outcome is to help the participant to get a better understanding of the whole process. <ref> Marketing Management Simulation Game [online]. [cit. 2016-01-24]. Available at: http://www.cesim.com/simulations/cesim-simbrand-marketing-management-simulation-game </ref>

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T22:08:39Z

Xtomp36: /* Possible costumer segmentation techniques */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

[[File:segment.jpg]]

==Usage segmentation==

Another rather simple approach of customer segmentation. There are two methods for usage segmentation either the customers are divided based on their weight of use or by time and place of usage. With the first method it is obvious that customers who buy more are more important to the business that the other ones. In here for example a “Pareto analysis” could be used to identify the top 20 % of most valuable customers. This method is normally used in business-to-business markets. The second method divides customers based on time and place. At different times customers may want different products available, so this method takes that into consideration. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Simulator and simulation game=
This part of the article describes two cases of customer segmentation usage, where simulation is a factor.

==Market Segmentation Simulator==
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

==Marketing Management Simulation Game==
This simulation game has some aspects of customer segmentation, so I find it worth mentioning in this article. It is developed by Cesim SimBrand and targets the marketing decision making process. It has 8 customer segments, 2 market areas and six different products. The game covers many others areas of marketing, than just customer segmentation, like: product life cycle management, positioning, after sales services, pricing, sales forecasting, competitor analysis and profitability. Its expected outcome is to help the participant to get a better understanding of the whole process. <ref> Marketing Management Simulation Game [online]. [cit. 2016-01-24]. Available at: http://www.cesim.com/simulations/cesim-simbrand-marketing-management-simulation-game </ref>

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

File:Segment.jpg

2016-01-24T22:08:05Z

Xtomp36:

Customer segmentation techniques

2016-01-24T22:04:07Z

Xtomp36:

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Usage segmentation==

Another rather simple approach of customer segmentation. There are two methods for usage segmentation either the customers are divided based on their weight of use or by time and place of usage. With the first method it is obvious that customers who buy more are more important to the business that the other ones. In here for example a “Pareto analysis” could be used to identify the top 20 % of most valuable customers. This method is normally used in business-to-business markets. The second method divides customers based on time and place. At different times customers may want different products available, so this method takes that into consideration. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Simulator and simulation game=
This part of the article describes two cases of customer segmentation usage, where simulation is a factor.

==Market Segmentation Simulator==
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

==Marketing Management Simulation Game==
This simulation game has some aspects of customer segmentation, so I find it worth mentioning in this article. It is developed by Cesim SimBrand and targets the marketing decision making process. It has 8 customer segments, 2 market areas and six different products. The game covers many others areas of marketing, than just customer segmentation, like: product life cycle management, positioning, after sales services, pricing, sales forecasting, competitor analysis and profitability. Its expected outcome is to help the participant to get a better understanding of the whole process. <ref> Marketing Management Simulation Game [online]. [cit. 2016-01-24]. Available at: http://www.cesim.com/simulations/cesim-simbrand-marketing-management-simulation-game </ref>

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:44:33Z

Xtomp36: /* Simulator and simulation game */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Simulator and simulation game=
This part of the article describes two cases of customer segmentation usage, where simulation is a factor.

==Market Segmentation Simulator==
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

==Marketing Management Simulation Game==
This simulation game has some aspects of customer segmentation, so I find it worth mentioning in this article. It is developed by Cesim SimBrand and targets the marketing decision making process. It has 8 customer segments, 2 market areas and six different products. The game covers many others areas of marketing, than just customer segmentation, like: product life cycle management, positioning, after sales services, pricing, sales forecasting, competitor analysis and profitability. Its expected outcome is to help the participant to get a better understanding of the whole process. <ref> Marketing Management Simulation Game [online]. [cit. 2016-01-24]. Available at: http://www.cesim.com/simulations/cesim-simbrand-marketing-management-simulation-game </ref>

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:44:10Z

Xtomp36: /* Simulators and simulation game */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Simulator and simulation game=
This part of the article shows two cases of customer segmentation usage, where simulation is a factor.

==Market Segmentation Simulator==
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

==Marketing Management Simulation Game==
This simulation game has some aspects of customer segmentation, so I find it worth mentioning in this article. It is developed by Cesim SimBrand and targets the marketing decision making process. It has 8 customer segments, 2 market areas and six different products. The game covers many others areas of marketing, than just customer segmentation, like: product life cycle management, positioning, after sales services, pricing, sales forecasting, competitor analysis and profitability. Its expected outcome is to help the participant to get a better understanding of the whole process. <ref> Marketing Management Simulation Game [online]. [cit. 2016-01-24]. Available at: http://www.cesim.com/simulations/cesim-simbrand-marketing-management-simulation-game </ref>

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:42:02Z

Xtomp36: /* Marketing Management Simulation Game */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Simulators and simulation game=
==Market Segmentation Simulator==
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

==Marketing Management Simulation Game==
This simulation game has some aspects of customer segmentation, so I find it worth mentioning in this article. It is developed by Cesim SimBrand and targets the marketing decision making process. It has 8 customer segments, 2 market areas and six different products. The game covers many others areas of marketing, than just customer segmentation, like: product life cycle management, positioning, after sales services, pricing, sales forecasting, competitor analysis and profitability. Its expected outcome is to help the participant to get a better understanding of the whole process. <ref> Marketing Management Simulation Game [online]. [cit. 2016-01-24]. Available at: http://www.cesim.com/simulations/cesim-simbrand-marketing-management-simulation-game </ref>

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:32:59Z

Xtomp36:

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Simulators and simulation game=
==Market Segmentation Simulator==
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

==Marketing Management Simulation Game==

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

WS 2015/2016

2016-01-24T21:30:51Z

Xtomp36: /* Papers */

Semestral papers from winter term 2015/2016. Please, put here links to the pages with your paper. First you need to have your [[Assignments WS 2015/2016|assignment approved]].

==Simulations==
--[[User:Xkrep33|Xkrep33]] ([[User talk:Xkrep33|talk]]) 15:35, 15 January 2016 (CET) [[Maze Solving Robot Simulation]]

--[[User:Dinara|Dinara]] ([[User talk:Dinara|talk]]) 22:30, 15 January 2016 (CET) [[Aircraft boarding methods]]

--[[User:OtakarTrunda|OtakarTrunda]] ([[User talk:OtakarTrunda|talk]]) 11:48, 16 January 2016 (CET) [[Ultimatum Game]]

--[[User:Xpokl18|Xpokl18]] ([[User talk:Xpokl18|talk]]) 06:51, 17 January 2016 (CET) [[RetrieverBreeder]]

--[[User:Martin.zima|Martin.zima]] ([[User talk:Martin.zima|talk]]) 10:28, 17 January 2016 (CET) [[Lane-merge optimization]]

--[[User:Xhudj17|Xhudj17]] ([[User talk:Xhudj17|talk]]) 11:25, 17 January 2016 (CET) [[Crossroad simulation]]

--[[User:xtomp36|xtomp36]] ([[User talk:xtomp36|talk]]) 15:33, 17 January 2016 (CET) [[Load-balancing]]

==Papers==
--[[User:Xkrep33|Xkrep33]] ([[User talk:Xkrep33|talk]]) 18:14, 22 January 2016 (CET) [[System Archetypes]]

--[[User:Xpokl18|Xpokl18]] ([[User talk:Xpokl18|talk]]) 07:44, 23 January 2016 (CET) [[PPC systems, Ad Rank]]

--[[User:Dinara|Dinara]] ([[User talk:Dinara|talk]]) 14:47, 23 January 2016 (CET) [[System Dynamics]]

--[[User:Knam00|Knam00]] ([[User talk:Knam00|talk]]) 17:11, 24 January 2016 (CET) [[Shifting the Burden Archetype]]

--[[User:xtomp36|xtomp36]] ([[User talk:xtomp36|talk]]) 22:29, 24 January 2016 (CET) [[Customer segmentation techniques]]

WS 2015/2016

2016-01-24T21:30:35Z

Xtomp36: /* Papers */

Semestral papers from winter term 2015/2016. Please, put here links to the pages with your paper. First you need to have your [[Assignments WS 2015/2016|assignment approved]].

==Simulations==
--[[User:Xkrep33|Xkrep33]] ([[User talk:Xkrep33|talk]]) 15:35, 15 January 2016 (CET) [[Maze Solving Robot Simulation]]

--[[User:Dinara|Dinara]] ([[User talk:Dinara|talk]]) 22:30, 15 January 2016 (CET) [[Aircraft boarding methods]]

--[[User:OtakarTrunda|OtakarTrunda]] ([[User talk:OtakarTrunda|talk]]) 11:48, 16 January 2016 (CET) [[Ultimatum Game]]

--[[User:Xpokl18|Xpokl18]] ([[User talk:Xpokl18|talk]]) 06:51, 17 January 2016 (CET) [[RetrieverBreeder]]

--[[User:Martin.zima|Martin.zima]] ([[User talk:Martin.zima|talk]]) 10:28, 17 January 2016 (CET) [[Lane-merge optimization]]

--[[User:Xhudj17|Xhudj17]] ([[User talk:Xhudj17|talk]]) 11:25, 17 January 2016 (CET) [[Crossroad simulation]]

--[[User:xtomp36|xtomp36]] ([[User talk:xtomp36|talk]]) 15:33, 17 January 2016 (CET) [[Load-balancing]]

==Papers==
--[[User:Xkrep33|Xkrep33]] ([[User talk:Xkrep33|talk]]) 18:14, 22 January 2016 (CET) [[System Archetypes]]

--[[User:Xpokl18|Xpokl18]] ([[User talk:Xpokl18|talk]]) 07:44, 23 January 2016 (CET) [[PPC systems, Ad Rank]]

--[[User:Dinara|Dinara]] ([[User talk:Dinara|talk]]) 14:47, 23 January 2016 (CET) [[System Dynamics]]

--[[User:Knam00|Knam00]] ([[User talk:Knam00|talk]]) 17:11, 24 January 2016 (CET) [[Shifting the Burden Archetype]]

--[[User:xtomp36|xtomp36]] ([[User talk:xtomp36|talk]]) 22:29, 24January 2016 (CET) [[Customer segmentation techniques]]

Customer segmentation techniques

2016-01-24T21:26:23Z

Xtomp36: /* Market Segmentation Simulator */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Market Segmentation Simulator=
There is a conjoint simulator web application, which gives the ability to predict the market share of new products and to measure the gain or loss in market share based on changes to existing products. this application uses the Conjoint Analysis described before in this article. So the user needs to define the attributes and levels, which are then used to create product concepts (profiles) by the simulator application. The goal of the application is to simulate the market share of the products to establish a baseline.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:20:25Z

Xtomp36:

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Market Segmentation Simulator=

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:15:29Z

Xtomp36: /* Customer segmentation in SimProcess */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.
Other than this it is easier to imagine the usage of customer segmentation in simulations. As the segmentation is already done and it is used only as input for the simulation.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:11:29Z

Xtomp36: /* Clustering approach */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]

Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:11:22Z

Xtomp36: /* Clustering approach */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

[[File:hi_cluster.jpg]]
Figure 1 - graphical example of hierarchical clustering <ref> Hierarchical clustering [online]. [cit. 2016-01-23]. Available at: http://people.revoledu.com/kardi/tutorial/Clustering/Numerical%20Example.htm </ref>

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

File:Hi cluster.jpg

2016-01-24T21:09:58Z

Xtomp36:

Customer segmentation techniques

2016-01-24T21:06:11Z

Xtomp36: /* Customer segmentation in general */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential. <ref> Customer segmentation definition [online]. [cit. 2016-01-24]. Available at: http://searchsalesforce.techtarget.com/definition/customer-segmentation </ref>

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences). <ref> Segment your customers [online]. [cit. 2016-01-24]. Available at: http://www.infoentrepreneurs.org/en/guides/segment-your-customers/ </ref>

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

Zde možná obrázek

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:03:53Z

Xtomp36: /* Customer segmentation in general */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

Zde možná obrázek

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:03:31Z

Xtomp36: /* A priory segmentation */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).
…
Steps how to deal with CS

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref>

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

Zde možná obrázek

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:02:39Z

Xtomp36: /* Needs based segmentation */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).
…
Steps how to deal with CS

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude.

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller. <ref> Market segmentation [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/segmentation.htm </ref> <ref> Conjoint analysis [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Conjoint/Conjoint_analysis.htm </ref>

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

Zde možná obrázek

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T21:00:55Z

Xtomp36: /* Qualitative approach */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).
…
Steps how to deal with CS

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude.

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller.

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable. <ref> Qualitative research [online]. [cit. 2016-01-23]. Available at: http://www.dobney.com/Research/qualitative_research.htm</ref>

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

Zde možná obrázek

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T20:59:28Z

Xtomp36: /* Resources */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).
…
Steps how to deal with CS

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude.

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller.

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable.

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

Zde možná obrázek

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

Customer segmentation techniques

2016-01-24T20:59:18Z

Xtomp36: /* Clustering approach */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).
…
Steps how to deal with CS

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude.

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller.

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable.

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>Approaches to Segmentation [online]. [cit. 2016-01-24]. Available at: http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

Zde možná obrázek

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf

Customer segmentation techniques

2016-01-24T20:57:46Z

Xtomp36: /* Clustering approach */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).
…
Steps how to deal with CS

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude.

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller.

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable.

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter. <ref>http://www.marketingdecisions.net/Arts-Notes/Segmentation1.html</ref>

Zde možná obrázek

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf

Customer segmentation techniques

2016-01-24T20:56:48Z

Xtomp36: /* Customer segmentation in SimProcess */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).
…
Steps how to deal with CS

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude.

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller.

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable.

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter.

Zde možná obrázek

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]

Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf

Customer segmentation techniques

2016-01-24T20:56:41Z

Xtomp36: /* Customer segmentation in SimProcess */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).
…
Steps how to deal with CS

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude.

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller.

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable.

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter.

Zde možná obrázek

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]
Figure 2 - Possible model for SimProcess <ref>Allocation of Marketing Resources to Optimize Customer Equity [online]. [cit. 2016-01-23]. Available at: http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf</ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf

Customer segmentation techniques

2016-01-24T20:55:32Z

Xtomp36: /* Customer segmentation in SimProcess */

*'''The Essay topic''': Customer segmentation techniques (article)
*'''Class''':[http://isis.vse.cz/katalog/syllabus.pl?predmet=95726;zpet=..%2Fpracoviste%2Fpredmety.pl%3Fid%3D64%2Clang%3Dcz;jazyk=1;lang=sk 4IT496 System Simulation] (WS 2015/2016)
*'''Author''': Bc. Patrik Tomášek (xtomp36)

=Introduction=
This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

=Customer segmentation in general=
What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).
…
Steps how to deal with CS

=Possible costumer segmentation techniques=

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

==A priory segmentation==

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude.

==Needs based segmentation==

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller.

==Qualitative approach==

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable.

==Clustering approach==

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter.

Zde možná obrázek

=Customer segmentation in SimProcess=

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

[[File:cs_model.jpg]]
Figure 2 <ref>Understanding ad position and Ad Rank. Google Support - AdWords Help [online]. [cit. 2016-01-23]. Available at: https://support.google.com/adwords/answer/1722122?hl=en </ref>

This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

=Summary=

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

=Resources=

http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf

Customer segmentation techniques

2016-01-24T20:53:06Z

Xtomp36:

Customer segmentation techniques

2016-01-24T20:52:34Z

Xtomp36: /* Customer segmentation in SimProcess */

Customer segmentation techniques

2016-01-24T20:52:25Z

Xtomp36: /* Customer segmentation in SimProcess */

File:Cs model.jpg

2016-01-24T20:52:04Z

Xtomp36:

Customer segmentation techniques

2016-01-24T20:51:22Z

Xtomp36:

Customer segmentation techniques

2016-01-24T20:49:51Z

Xtomp36:

Customer segmentation techniques

2016-01-24T20:48:30Z

Xtomp36:

This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).
…
Steps how to deal with CS

'''Possible costumer segmentation techniques'''

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

'''A priory segmentation'''

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude.

'''Needs based segmentation'''

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller.

'''Qualitative approach'''

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable.

'''Clustering approach'''

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, ''partition clustering'' and ''hierarchical clustering''.

With ''partition clustering'' it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter.

Zde možná obrázek

'''Customer segmentation in SimProcess'''

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf
This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

'''Summary'''

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

Customer segmentation techniques

2016-01-24T20:47:32Z

Xtomp36: Customer segmentation techniques

This article briefly describes the basics of customer segmentation, what it means and what its benefits are. Furthermore some of the most common techniques are also mentioned and described.

What is customer segmentation in general? It is a sort of practice which divides customers in smaller groups based on multiple characteristics and gives the business a better understanding of its customers, which has a great value in B2C relationship. Some of the other benefits could be identification of least profitable customer group, improvement of customer service, data for making marketing decisions (which group to target) or avoiding unprofitable markets. The main goal of this whole process should be to maximize the value of each single customer. There could be other goals of customer segmentation, however it rather depends on the particular business and its preferences. Examples of common segmentation objectives include development of new product, differentiated customer servicing or targeting prospects with the highest profit potential.

Of course the first expectation to make the customer segmentation possible is that the customer base is dividable. That is determined by the data, which are collected by the particular business, wanting to segment its customer base. The input keys are used to differentiate customers could be demographical (like age, gender, income), psychographic (lifestyle), geographic (geo location – where the customer lives) or behavioral (spending habits, product preferences).
…
Steps how to deal with CS

'''Possible costumer segmentation techniques'''

There are many possible customer segmentation techniques, I will briefly explain some of the more common ones. The type of segmentation used will very on a lot of factors (goals, costs, etc.).

'''A priory segmentation'''

This is one of the most simple approaches, where the market is divided according to already existing segments (that is why a priory – “pre-existing”) such as gender or age. In some businesses this approach of segmentation could be sufficient, for example the technology sector has a strong relationship between age and usage or product preferences. In other sectors it might be more difficult to segment the customer base using this method. Without any doubt sing this method is better than pure mass marketing, however it is still quite crude.

'''Needs based segmentation'''

This approach mostly uses co called “Conjoint Analysis”, which is and advanced research technique (it also known as Discrete Choice Estimation). It gives the customer choices and then analysis why they made them. The output of such analysis is a measurement of utility. Firstly, with conjoint analysis a product or service is divided into its constituent parts. Than the possible combinations of these parts are tested to find which combinations are preferred by the customers. Furthermore each part may have multiple attributes (for example memory has a size and frequency). These attributes are defined in levels a computer memory can for example have 4 GB or 8 GB and operate on 1600 or 2133 MHz. Attributes and levels are used to define products and one of the first steps of conjoint analysis is to define a set of product profiles, which represent choices for customers. The number of possible product profiles rapidly increases with each additional attribute, therefor it is important to find balance between the number of attributes and complexity of customer choice, in order to get quality results. After this a range of statistical tools can be used to analyze which items customers choose or prefer from the product profiles. With need based segmentation the specific individual needs are identified and with the above approach they can be used to find other products, which meet the same customer requirements. Such products can then be offered to a particular customer by the seller.

'''Qualitative approach'''

An approach about why and what for are (in this case customers) behaving the way they are. Usually the outcome of this method is not enough to base a statistic on, it rather focuses on getting several different surveying. The aim is to get the best picture about the market and the customers so that one could spot the gaps, the similarities and the differences and exploit them. Focus groups or individual depth interviews are very often used to clarify the behavior and attitude of people. Sometimes it might be a good idea to use another technique such as conflict groups, triads or paired interviews. Outcome of these techniques is researcher (or a moderator) dependent, meaning he must be experienced in order to get the correct information out of the interview. It may happen that the participants may not be willing to speak their mind freely and will just repeat what they believe to be right. It is safe to say that a mind probe would make this approach a lot easier, unluckily this technology we do not possess. The discussion itself usually starts with a really broad term or subject and is later narrowed down by the moderator without putting too much pressure on the interviewed meanwhile getting an idea about how their beliefs and views of the subject changes throughout the session. The most valued facts about these participants are in the end their mood and their spontaneous reactions. Taking in count any information that is artificially created by the interviewees to cover up their true opinion is lowering the value of the session and of the finding itself. Though describing this technique defies its original purpose to be original every single time and it is contentious whether or not the results are valuable.

'''Clustering approach'''

This is one of the more advanced approaches, because it can be used to split customers based on the way they think or feel, rather than only who they are. This method start with individuals and forms segments with them based on similarities. At first it is necessary to choose one or more criteria, which will be used to assess the similarity between individuals (as already states it is common to use behavioral data with clustering). A good way to search for input data might be in surveys, which often focus on customer attitude with the business or product. If multiple criteria are in place, it might be necessary to standardize them so that they have comparable weight. Secondly it is important to choose clustering approach that will be used. There are two commonly applied approaches to clustering, Partition clustering and hierarchical clustering.
With partition clustering it is decided how many clusters will be (or better should be, because the result needn’t be possible) formed at the beginning. There are multiple methods of computation. One of the basic ones lies in specifying the “seed” observations for each one of the desired clusters. Than calculate the distance from each of the remaining observations to each seed and assign each observation to the nearest seed to form a set of clusters. This process can be repeated to reach greater homogeneity.

Zde možná obrázek

''Hierarchical clustering'' is the second common approach, is based on assumption that closer objects are more related than the further ones. The algorithm begins by finding a most similar pair of observations in terms of previously defined criteria and joins them to form a group. The next part is to find another pair, which can be again joined together or joined to a previously formed group if the distance is shorter.

Zde možná obrázek

'''Customer segmentation in SimProcess'''

It is rather hard to imagine the usage of customer segmentation in a program like SimProcess or Netlogo. However in SimProcess it might be possible to model the customer behavior and segment them based on predefined actions.

http://www1.unisg.ch/www/edis.nsf/SysLkpByIdentifier/3038/$FILE/dis3038.pdf
This model could be used as a bases for possible design of a SimProcess simulation. Of course a generation of customers’ needs to be added. After that this could lead to finding out how many customers of different groups the business could have.

'''Summary'''

This article has introduced to the reader the topic of customer segmentation, its basics, usage and some of the possible techniques, which are used to segment customers. It is clear that most of the segmentation techniques aren’t simple and are impossible without proper solution, however the business value is undeniable. Of course the technique used for segmentation will depend on the business needs.

User talk:Xtomp36

2016-01-24T20:44:35Z

Xtomp36: /* Customer segmentation techniques */ new section

== Customer segmentation techniques ==

test

File:Dos mittigation.jpg

2016-01-17T22:57:08Z

Xtomp36: Xtomp36 uploaded a new version of File:Dos mittigation.jpg

Assignments WS 2015/2016

2016-01-17T22:54:11Z

Xtomp36: /* Hosting load-balancing simulation */

{{DISPLAYTITLE:Assignments WS 2015/2016}}

{{Ambox
| text = <div>
Please, put here your assignments. Do not forget to sign them. You can use <nowiki>~~~~</nowiki> (four tildas) for an automatic signature. Use Show preview in order to check the result before your final sumbition.
</div>
}}

{{Ambox
| text = <div>
Please, strive to formulate your assignment carefully. We expect an adequate effort to formulate the assignment as it is your semestral paper. Do not forget that your main goal is a research paper. It means your simulation model must generate the results that are specific, measurable and verifiable. Think twice how you will develop your model, which entities you will use, draw a model diagram, consider what you will measure. No sooner than when you have a good idea about the model, submit your assignment. And of course, read [[How to deal with the simulation assignment|How to deal with the simulation assignment]].
</div>
}}

{{Ambox
| type = content
| text = <div>
In order to avoid possible confusion, please, check if you have added '''approved''' in bold somewhere in our comment under your submission. If there is no '''approved''', it means the assignment was not approved yet.
</div>
}}

= Assignments =

[[User:Oleg.Svatos|Oleg.Svatos]] ([[User talk:Oleg.Svatos|talk]]) 15:35, 29 November 2015 (CET) Just to warn you in advance - topics like roulette, lottery and other simulations based purely on uniform randomnes '''will not be accepted''' as Monte Carlo simulation.

--[[User:Dinara|Dinara]] ([[User talk:Dinara|talk]]) 03:32, 11 December 2015 (CET)
I have changed my topic to Chemicals manufacturing process simulation

==mand01: Chemicals manufacturing process simulation==
--[[User:Dinara|Dinara]] ([[User talk:Dinara|talk]]) 21:24, 14 December 2015 (CET)
===Introduction===
The simulation shows how pharmaceutical companies' manufacturing process is held. In this case manufacturing process is a chemical batch production process which has shared resources. There are two basic systems included: Ordering system and Production systems. The first one generates orders of chemicals to be produced with different demands on a production. As this simulation is for educational purposes, I have decided to simplify the model to 2 types of drugs produced. The production system is assigning reactors to a particular order and processing it. We have shared resources and one of the main shared resource is a reactor. One reactor processing one order. Once an order is received we align each per order and then production begins, which include several steps as adding water, heating water, adding particles, stiring and draining. After all production steps we release the reactor for the next order.

===Problem definition===
This model explores how to solve resource allocation problem.

We need to process 100 orders (60 orders for drug A and 40 for drug B) in a month. Reactors works 12 hours a day. We obtain 5 reactors, we can add water to two reactors simultaneously and heat two reactors at a time.

'''Drug A''': adding water 2 hours, heating 3 hours, adding particles 1 hour, stiring 3 hours and draining 3 hours

'''Drug B''': adding water 1 hour, heating 2 hours, adding paticles 30 min, stiring 5 hours and draining 3 hours

Do we need an additional reactor to acomplish this task?

===Method===
Tool to be used - ''Simprocess''

As stated [[How_to_deal_with_the_simulation_assignment|here]], for discrete event simulations a real problem from real business with real data is mandatory.

== mand01: Aircraft boarding methods ==
--[[User:Dinara|Dinara]] ([[User talk:Dinara|talk]]) 04:41, 16 December 2015 (CET)
===Introduction ===
This simulation explores different methods of aircraft boardings. Airports serves thousands of people every day and one of the main problems they meet is the efficiant boarding of passengers.

Exists several methods of boarding:

- random (which is widely used, the most common): people just taking their seats without any specific order

- WMA (Windows-Middle-Aisle) - boarding is divided into three groups: seats near the windows, middle seats and aisle seats. The first boarding group is windows then middle and finally the aisle group.

- Back-front WMA: we divide groups in the same way as in WMA, but the only difference, that we sort passengers in decreasing order. For each group boadring starts from back of the plane.

=== Goal ===
To simulate different boarding scenarious and try to find out the optimal one.

=== Method ===
Environment: NetLogo

Agents: passengers (row number, seat number, isSitting, isMoving, luggage, luggageLoadTime, orderingNumber)

Simulation ends when all passengers are sitting (isSitting == true). Every agent is independent, random number of passengers are with luggage, luggage has the same size.

For simulation we can chose the boarding method, probability of passengers with luggage. Other parameters are fixed. Plane: 18 rows x 6 seats (3 per each side).

'''Approved''' [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 19:47, 16 December 2015 (CET)

[[xpokl18]]

==knam00:Heating system simulation==
===Intro===
<div>
Most households in the Czech Republic have mains gas central heating. This is a so-called ‘wet system’, which means a gas-fired boiler heats water to provide central heating through radiators and hot water through the taps in your home.

Some houses that aren’t connected to the gas network can use electrical heating or liquid petroleum gas (LPG) or heating oil, which work in a similar way to gas central heating, although LPG and oil are delivered by road and stored in a tank, which you may have to buy or rent from your supplier.

Gas is a highly efficient fuel, so you get a good return on every unit of energy. Modern condensing boilers, which use hot flue gases that are wasted in a standard boiler, have very high efficiency. Some are now 90% or more efficient.
</div>
<div>
===Simulation design===
====Layout====
[http://imageshack.com/a/img910/6936/OIUdr1.png Preview Heating system]
====People====
There are 6 people using this flat. Each person periodically comes and leaves.
Pair 1 comes on monday and leaves on wednesday
Pair 2 comes on monday and leaves on friday
There are also 2 boys sharing another room - The first one stays here for the whole week and the second one leaves on saturday and comes on monday.

====Gas boiler====
We own 25kW condensing boiler with 95% efficiency.
Boiler is controlled with thermostat.

====Thermostat====
Thermostat is placed in the living room - it is set to 22 degrees Celsius from 7:00 to 23:00 and to 19 degrees Celsius from 23:00 to 7:00.

====Rooms====
There are 3 separated rooms used privately, one living room, kitchen, bathroom, storage room and toilet.
He have high ceilings, so volume of rooms is really big.

Private rooms - 84mˆ3 x 3
Living room - 168mˆ3
Kitchen - 84mˆ3
Bathroom - 21mˆ3
Storage room - 21mˆ3
Toilet - 21mˆ3

====Radiators====
We have old radiators made from cast-iron. Each part of this radiator is 600mm high and 200mm deep, its surface is 0,31mˆ2, its capacity is 1,7L.
Each room (except storage room) has its own radiator.
In total, there are 5 radiators with 30 parts and 2 radiators with 15 parts.

====Heating system====
Consists of radiators and tubes. Total amount of water inside heating system is 255L (5 big radiators) + 51L (2 smaller radiators) + 61L (Amount of water inside tubes - 108m of tubes in total) = 367L.

====System dynamics====
Water inside heating system gets heated - it takes 4181J to heat up 1L of water to 1 degree Celsius. Then radiators are heated (416J for 1kg of cast-iron to heat-up to 1 degree Celsius), then air is heated according to Stefan-Boltzman law. Heat is distributed among the flat. According to the people in the flat, radiators are turned on or off. Thermostat is inside the living room a living room gets 30% of heat from rooms connected to it. Thermostat stops heating system when requirement are satisfied. Accumulated heat stays in heating system and slowly degrades. I would like to measure amount of gas used for heating system for the period 4 months.

The system i describe doesn't count with using hot water for shower.
</div>
<div>
I would like to ask you for software recommendation - not sure if i should use simprocess or vensim.
</div>
<div>
Martin Knapovský, knam00@vse.cz
</div>

:This could be an interesting topic, but please, finish the assignment. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 02:44, 13 December 2015 (CET)

::From my point of view, systems' dynamics could be a good one. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 21:49, 15 December 2015 (CET)
::: System dynamics seems fine with me.'''Approved''' [[User:Oleg.Svatos|Oleg.Svatos]] ([[User talk:Oleg.Svatos|talk]]) 08:05, 16 December 2015 (CET)

== Christmas market ==
<div>
I´m thinking about simulation of Christmas market, on some square as place for vendors. There will be two kind of vendors – with refreshments and with decorations. There will be two entities – vendors and customers.
In first round I want place there some number of customers (for example twenty) and some number of vendors.
Each customer has to buy at least one decoration as a present for somebody. In next rounds each customer can decide if he want to leave a market, buy other present or buy something to eat/drink. Each customer can buy no more than five presents.
Every minute will come five new customers.
There will be a temperature as global variable.
If the temperature is too low, customer will decide, if he/she wants to go home or buy something to eat/drink to get warm.
If customer is on a market for so long, he/she needs to buy something to eat or drink either.
Because there will be more than one vendor of each kind of goods, customer will preference vendor by the distance.
</div><div>
My goal is to find an optimal number of vendors and its structure based on the given temperature.
</div><div>
For my simulation I decide to use NetLogo.
</div>
<div>--[[User:Xhejk15|Xhejk15]] ([[User talk:Xhejk15|talk]]) 20:58, 13 December 2015 (CET)</div>
:This could be interesting if you would use real data. Otherwise it is kinda academic exercise. And by the way, it seems to be rather a discrete event simulation than an agent model. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 21:52, 15 December 2015 (CET)
<div>Ok, look at my second assignment below please, in that case I can use real data.</div>--[[User:Xhejk15|Xhejk15]] ([[User talk:Xhejk15|talk]]) 19:42, 20 December 2015 (CET)

== Fastfood restaurant ==
<div>
I want to simulate service in fastfood restaurant during one day. For this simulation I have some real data from my friend who works there.
</div><div>
Opening hours: 06:00 – 02:00
</div><div>
Capacity of restaurant: 106
</div><div>
Number of cashier's desk: 5
</div><div>
I can get some real data about how much employees are in the kitchen/at cashier's desk and number of transactions for each hour in day.
</div><div>
During lunch 35% of customers at average take their food away. On the other side during night time 65% of customers at average take their food away.
</div><div>
One order take at average 1 minute to finish. If there is some special order and it´s not completed from the kitchen yet, it takes kitchen employees to complete this order about 3 minutes in average. Customers with special orders don´t wait in queue, but they sit down and wait for it at the table. So, employee at cashier desk can service next customer.
</div><div>
For my simulation I would like to use Simprocess.
</div><div>
Goal of my simulation is to see if there is enough cashier´s desks, capacity of restaurant and employees.
</div>
[[User:Xhejk15|Xhejk15]] ([[User talk:Xhejk15|talk]]) 19:42, 20 December 2015 (CET)

::That's pity you didn't remind me. Unfortunately, this topic is too simple. It is something like we solved as a class example. Please, try something more sophisticated.

::::So, can I expand this assignment somehow? I don´t really have an idea. What´s your expectations? Thank you. [[User:Xhejk15|Xhejk15]] ([[User talk:Xhejk15|talk]]) 15:24, 2 January 2016 (CET)

:::::It is hard to say in general, what level of complexity is expected. What you have proposed is very similar to the task we did during our classes. For seminal paper, more complicated problem is needed. The appropriate level of complexity could be perhaps apparent from the approved assignments. In the case of your problem, in order to make it meaningful, you could e.g. simulate the whole day with real customer flow. It means much more complex data input reflecting changes during the day. It is perfect that you have someone who can provide real data. Try to get the records from cash registers (just times of transactions are enough) and load them into Simprocess. Next, if you obtain real cost structure, you can build cost model. I would say in the case of orders, it depends on the kind of ordered meal. And what about production part of the whole thing (kitchen) - does it have any influence on the capacities of the restaurant? Etc. Of course, you can also try to suggest something completely different. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 02:35, 3 January 2016 (CET)

:::::::I was talking to my friend and he told me that he unfortunately cannot provide me times of transactions, because there is some sensitive data also. But I can have number of transactions during each hour of whole day. So if I will simulate whole day it would be better? He cannot provide me cost structure neither. Maybe I can add to simulation how customers wait for their specials orders. [[User:Xhejk15|Xhejk15]] ([[User talk:Xhejk15|talk]]) 19:29, 5 January 2016 (CET)

::::::::I am afraid that in such a situation, the assignment is so artificial that I don't consider it appropriate. Just to explain - I generally don't expect complicated assignments, however, what you have proposed, is really simple and too much resembling the tasks that have been solved during our lessons. That's why I asked for real data in order to make it more meaningful. If they are not available, I would prefer another assignment. Really, what you have proposed could be solved during one afternoon. This is not what I would expect from a term paper. Thanks for your understanding. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 22:41, 5 January 2016 (CET)

:::::::::: Ok, so what about simulate something different - like traffic with semafors vs. roundabout vs. classic crossroad with preference. And then see what is the best? Could I use Netlogo for simulation like this? [[User:Xhejk15|Xhejk15]] ([[User talk:Xhejk15|talk]]) 00:14, 7 January 2016 (CET)

::::::::::: This could be a good idea, but it is necessary to elaborate the assignment into a greater details. As we are terribly late with this, I strongly recommend to let me know on my email that you have updated this. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 23:39, 15 January 2016 (CET)

== xsedd07: assignment ==
Conference and meeting center layout in '''NetLogo'''

This simulation would display how the crowds (and individuals) would behave within a conference center. The building would have a set plan, what would be changing could be the exits, the information desks, the buffets, the bars, the exhibitions, the lectures, the meetings, the benches and the free space. The visitors and participants would move around based on a schedule and free will, at this point I am not sure to which extent is it wise to design an AI which would control each person as an individual. Hopefully such AI can be made with NetLogo spending just reasonable effort. The simulation extent (regarding total amount of people, rooms...) will be adjusted so that it can be easily overviewed but not yet too simplified.

The layout will be transformed multiple times, its efficiency will be judged by density of the crowd and queues. Individual buffets, bars e.t.c. should not be abandoned while the others are flooded. Simulation will include staff and toilets to make the environment look more real. At the moment I am open to implementing a possible terrorist attack, which could also be an alternative for this assignment.

The simulation should track the number of people in lecture halls, in queues and in any otherwise interesting state such as boredom or stress.
: I would make it simpler. Don't bother simulating all the day, simulate just as people are coming for the particular event/presentation. Simulate how crowd are assembling at doors, toilets, how people avoiding each other to get their seats. Etc. Just the time let's say +-10 minutes around the beginning of the event. Of course, you need to simulate relative intelligent movement of agents. If you agree, then it is '''approved'''. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 21:58, 15 December 2015 (CET)

== xtomp36: assigment ==
===Hosting load-balancing simulation===

A hosting company with it’s own infrastructure is using so called "load balancing" to distribute the overall load between multiple servers (hardware) and “high-availability” to minimize service down-time.

'''Software used''' 
- SIMPROCESS

'''Possible hosting services''' 
- Web hosting 
- VPS 
- Communication server (TS 3)

The simulation should also consider critical situation like off-line server or unavailability (for example due to D-DOS attack) of the entire server location (datacenter). In with case the traffic should be re-routed to another location (there are two hosting locations). 
I am going to use real data from my own hosting environment.

'''Variables''' 
- Number of web servers 
- Number of storage servers 
- Presence of anti-dos device 
- Load-balancers and DB server 

'''Other''' 
In addition there are other devices necessary to enable LB and HA, like a switch. In case of HA enabled there must be at least two same switches at one time to achieve redundancy.

'''Approved'''. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 21:59, 15 December 2015 (CET)

== Predator-Prey simulation ==
Few weeks ago a came across [http://www.natureworldnews.com/articles/6296/20140308/deer-overpopulation-threat-forest-growth-researchers.htm an interesting article] about deer overpopulation and how this is slowing down forest succession or natural establishment. 'The study was conducted on Cornell land near Freese Road in Ithaca,' as stated in the article. 
I would like to make a simulation of the predator-prey concept on this topic. We would be able to simulate the growth of deer, wolf population, the state of the 2 types of vegetation in the forest and how many deers and wolves would be optimal to keep the original forest vegetation from disappearing. 
Basically there will be total of 4 entities - 2 agents and 2 patch types: 
Deer - agent, consuming the native forest flora, 
Wolves - agent, the predator specie that hunts the deer specie, 
Forest vegetation - native forest flora that deer specie prefers to consume, 
Foreign vegetation - the more the deers eat the native forest vegetation the more this foreign flora thrives. 

NetLogo will be used for this simulation. I understand it might be a little more complex because this simulation would not be simulating only the deer population but the predator (wolves) as well. It would also simulate the interactions between deer and the native forest flora, wolves hunting deers and the foreign flora taking over the forest until there is little or none of the original forest vegetation. 
There will be some global variables such as the initial number of deers and wolves; deer, wolf, forest and foreign flora reproduction rates and rules of interaction for these entities.

--[[User:Xnovs00|Xnovs00]] ([[User talk:Xnovs00|talk]]) 21:57, 13 December 2015 (CET)

There is already so many predator-prey simulations (e.g. in Netlogo itself is at least one or two) that I am a bit reluctant to approve another one. Or please, let me know what is different on your particular case? 22:03, 15 December 2015 (CET)

== xhudj17: assigment ==
===Crossroad simulation===

This simulation should simulate trafic (cars and trams) and pedestrians on crossroad of roads Sokolovska, Jecna and Legerova located on I.P.Pavlova square in Prague.

'''Software used''' 
- NetLogo

'''Goal''' 

The simulation should try to optimize the trafic lights to reach lowest waiting times possible and to secure safety and fluency of the trafic.

'''Variables''' 
- Propability of incoming cars from each direction 
- Probability of incoming pedesterians 
- Probability of incoming trams 
- Light intervals 

'''Situation desctription''' 
The crossroad is combining 3 main streets and 3 smaller streets with several pedestrians crossing and two tram crossing. There are 2 separate trafic lights to controll the traffic.

Jan Hudecek, XHUDJ17, 14/12/2015 13:50

'''Approved.''' [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 22:04, 15 December 2015 (CET)

== xzimm00: Smart car traffic optimization ==

The model presented by this paper is simulating the situation of car multi-lane merging.

=== Goal ===

The primary goal of the simulation is to create a plausible real-world model of car traffic jams caused by agents operating in an inefficient way and environment that does not make the best of today's technologies that could help control the traffic flow, and to measure the possible improvement if smarter systems were employed.
The author takes into account the inefficiencies of the agents (human error...) and compares this with a (possibly better) solution using automated driver agents, always utilizing an (ideally) optimal (precomputed) traffic flow.

=== Variables ===
* Car inflow (cars/minute, distributed across the lanes)
* Number of lanes (starting and final, after the lane merge)
* Car speed (varying across the different road segments - i.e. max speed may be limited after the merge because of a traffic obstruction, for example a car accident or roadworks)
* Agents' merging strategy preference (i.e. late vs early merge)
* Agent inefficiency factor (reactions time, premature slowing down, wrong merging strategy used, varying speed, overcautiousness...)
* Amount of kept safety factor (for both human-based and automatic computer-based agent driven cars, so called 'defensive driving', expecting a failure of the others - i.e. mainly the distance kept between the cars)

=== Methods used ===
* Software used: NetLogo

=== Expected results ===
Based on the results of the simulation, it should be possible to measure the possible improvement (%) if automated car-driving agents were employed in place of humans in different critical situations of lane mergers.

(Martin Zima, xzimm00)
--[[User:Martin.zima|Martin.zima]] ([[User talk:Martin.zima|talk]]) 18:31, 14 December 2015 (CET)

'''Approved.''' [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 22:06, 15 December 2015 (CET)

== Maze Solving Robot Simulation ==
=== Assignment (xkrep33) ===
==== Goal ====
The primary goal is to create an autonomous robot, which is able to find way out of the maze. The robot will be programmed and his movements cannot be interfered after he starts. The robot must be driven by a non-trivial algorithm that will include randomly generated numbers. The robot must be able to find a way out of maze in a finite time. It means he should not get into an endless cycle of no return.

Secondary objective will be to determine what is the difference between a primitive robot and the robot uses a smarter algorithm. Primitive robot is e.g. such a robot who moves only straight forward and if he encounters an obstacle turns left.
==== Software ====
For simulation will be used NetLogo software (2D version).
==== Autonomous robot ====
Robot will be able to move up, down, left and right (viewed from above). He should be able to find the way out without any intervention. Robot will be able to move in any environment (maze) where exists at least one posible way out.
==== Environment (maze) ====
Environment will be represented by the World in NetLogo. The World will consists of black, grey and green patches. Black patches will represents the path where robot can move. The grey patches will represent walls, robot cannot enter them and finally the green patches will be the door out of the maze.

More than one enviroment will be provided. Environments will vary to prove that the robot is able to avoid an infinite loop situation.

--[[User:Xkrep33|Xkrep33]] ([[User talk:Xkrep33|talk]]) 20:46, 14 December 2015 (CET)

Sounds interesting. '''Approved.''' [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 22:08, 15 December 2015 (CET)

== Simulation of Aqua park ==
=== Assignment (xjanj58) ===
==== Description ====

The purpose of this simulation will be representation of Aqua park with indoor and outdoor pool. The swimmers will have random preferences and they will go to swim in one of two pools. With the pools gradually filling with swimmers it will have negative effect on motivation of each swimmer and it will show change of their preferences. Based on weather and temperature swimmers will also change their preferences between outdoor and indoor pool. Next parameter could be water purity, with high level of contamination produced by swimmers their motivation will lower and on the contrary.

==== Goals ====
Simulation will be set for period of one calendar year and at the end we can compare, if its profitable for Aquapark to have both pools open for a whole year. It will be cheaper to run outdoor pool, but bad weather will have much higher impact on outside swimmers. Next goal will be confirmation of assumption, that in summer people prefer outdoor pool and on the contrary in winter indoor pool.

==== Used software====
NetLogo

--[[User:xjanj58|xjanj58]] ([[User talk:xjanj58|talk]]) 22:36, 14 December 2015 (CET)

Sounds good, however you will probably struggle with soft parameters like "the willingness to swim in dirty water". How do you plan to deal with it? [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 22:23, 15 December 2015 (CET)

--- How about simplification in a way, that agents will produce "dirtyness" and there will be fix line, after which pool will be so dirty that nobody wants to swim in it. Indoor pool will have regular cleaning maintaince wich will lover level of dirtiness, but outside pool will only be cleaned after crossing this limit. It will take some time, lets say day or two to clean it, and then swimmers could go in again. [[User:xjanj58|xjanj58]] ([[User talk:xjanj58|talk]]) 9:56, 16 December 2015 (CET)

:: But, again, how do you turn this into specific numbers? How you will measure it? How do you know how quickly a pool gets dirty? And mainly, how will you measure the feeling of people about the quality of water. Those soft parameters are always a pain. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 19:52, 16 December 2015 (CET)
::--- I was thinking that agents will have parametr called e.g. "mood", and it will be interval between 0 and 10. Then for every new swimmer in the pool will this parametr go down and when it hit 0, then this particular agent will exit the pool. For dirtines, i was thinking thet pool will have some limit, for example 100, and every fifth move of the agent in the pool will fill this parametr by 1. When it reaches 100, the pool will be dirty. [[User:xjanj58|xjanj58]] ([[User talk:xjanj58|talk]]) 20:17, 16 December 2015 (CET)

::: Yes, I understand. This is exactly what I call "an academic exercise". How do you know that 0-10 is a correct interval to measure mood? How do you know that this is a linear function? Why it should go down by 1 (and not by 3 or 5 or 0.5) and why it should be the same for everyone? The same for the quality of water. Sure, you can make your model this way. But, will it be still in touch with reality? And if not, what's the point then? [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 21:01, 16 December 2015 (CET)

:: Ok,so how to add at least some randomnes in to model by include the random multiplier in agents formula.That way changes woudnt be constant and it can nicely represent diferences between human decisionmaking.
[[User:xjanj58|xjanj58]] ([[User talk:xjanj58|talk]]) 21:55, December 2015
(CET)

:::Try to find something that will not be dependent on such soft variables. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 13:17, 2 January 2016 (CET)

== Otakar Trunda: The Ultimatum Game ==

Agent-based simulation to analyze the socio-economical experiment known as The Ultimatum Game.

=== Game rules ===

The rules can be found at [https://en.wikipedia.org/wiki/Ultimatum_game]. During the game two players interact in order to divide certain amount of money (e.g. $1000 000) between them. At the beginning, the first player proposes the ratio of division (like $900 000 for himself and $100 000 for the other player, or any other ratio). Then the second player has two choices - he can either accept - in which case the players receive the agreed amounts of money, or he can reject - in which case no player receive any money.

=== Goals ===

There are several questions that can be studied about the Ultimatum Game. For example:

* how would people behave? (assuming the players don't know each other)
* what amount of money should the first player propose to the second? (to maximize his own profit)
* what amount of money should the second player accept?
* is it rational for the second player to reject any proposal? Why?
* if the game is played repeatedly and different players use different strategies, what strategy maximizes agents' long-run profit?
* if agents' strategies can evolve in time, can altruistic strategies emerge?
* are there evolutionary stable strategies? Which ones?

=== Method ===

In the system there will be agents representing players, each of them will play according to his strategy. The strategy of the player tells how much he will offer (should he be the first player in the game) and what is the least amount he will accept (should he be the second player in the game). In each simulation step, two agents will be selected randomly and play the game. Their score will be adjusted by the result of the game (i.e. if the second agent accepts, they will gain money according to the proposal, otherwise the score remains unchanged). Agents' successfulness is determined by their score.

After a large number of simulation steps (like 1000 000 of more), we should be able to determine how the strategy affects the final score of agents (i.e. what strategies are good).

=== Evolution of strategies ===

The system can be modified such that agents' strategies will change during the simulation to emulate biological evolution. In this case, successful agents will multiply in the system (i.e. more agents will use good strategies) and non-successful agents will die-out. During the reproduction of agents, their strategies will be slightly modified - the offspring will use a slightly different strategy then the parent. This way, the system should evolve to some kind of equilibrium where either all agents use very similar strategy (static equilibrium), or different strategies survive to compete with each other (dynamic equilibrium).

After a large number of simulation steps, we should be able to observe whether altruistic strategies emerge (strategies, that offer "fair" amounts and reject small offers) and determine the existence of stable strategies (strategies that can't be taken advantage of by "parasitic" strategies).

=== Software ===

Tailor-made in C# using Visual Studio.

[[User:Trunda|Trunda]] ([[User talk:Trunda|talk]]) 17:35, 15 December 2015 (CET)

The problem is well known. However, you will probably struggle with simulating irrationality and some other particular problems. The model will tend to be either pretty complex, almost hard to manage or too trivial. I would think twice and explain how exactly you will e.g. deal with the particular strategies.
Also, why do you want to use C# instead of a certain agent framework? [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 22:39, 15 December 2015 (CET)

I guess it is a well-known problem, but it seems interesting and I don't know of any similar problem that wouldn't be already well known. Can you perhaps recommend some?
How to deal with strategies: the strategy will be a pair of real numbers, lets denote them A and B from <0,1>. A will denote how much the strategy offers (e.g. A = 0.4 means that the player will offer 40% of the total amount to the second player) and B denotes how much at least the strategy accepts (e.g. B = 0.2 means that the player will reject any offer smaller than 20% of the amount and accept offers that are 20% or larger). Then the evolution can be done by a simple genetic algorithm on these pairs, fitness of the agent would be determined by its score.
I hope this setting is not too trivial, it can be further extended by adding a "space dimension" meaning that agents will only play the game with agents that are near them and during the reproduction the offspring will also be created near the parent. In this case there should emerge localized groups of agents using the same strategy, some groups should successively expand, others diminish.
I could use NetLogo, but I think it will be easier to do in C#. For example I would need to visualize the performance of the strategies - the score based on numbers A and B (mentioned earlier). This can be done in a form of a 3D graph, which I can create in C# easily using the proper library. I will also need to compute various statistics of the agents' scores to properly interpret the results (like moments and p-values) which again can be done easily using proper C# framework. It might be able to do these in NetLogo as well, but that would take me much more time :) [[User:Trunda|Trunda]] ([[User talk:Trunda|talk]]) 16:35, 16 December 2015 (CET)
::Good. According to what you have mentioned, you will also need to deal with irrationality. How you will do that? [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 19:56, 16 December 2015 (CET)
::::It depends on what do we call irrationality. I would like to think that any observed strategy would be "rational" in a sense that it evolved by maximizing the long-term profit. If an altruistic strategy evolves, then it means that our notion of rationality is too narrow and short-sighted and from the long-term perspective, altruism is rational. If it doesn't evolve on its own, we can at least set the altruistic strategy by hand and see if it survives in the environment (i.e. if it is stable). Furthermore, the successfulness of the strategy depends on the strategies used by other players so it is difficult to call a strategy rational or irrational without knowing the environment (strategies of its competitors). [[User:Trunda|Trunda]] ([[User talk:Trunda|talk]]) 21:48, 16 December 2015 (CET)

::::: '''Approved'''. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 13:19, 2 January 2016 (CET)

== ZCH webshop simulation ==
WHAT WILL I SIMULATE

I sell DVD movies on my own web shop on AUKRO. I sell DVD titles on a daily basis. The purchase price is CZK 20,- and the DVDs are sold for 35,- CZK (in average).From my sales report I deduced that the demand has probability distribution from 100 to 150 pieces. When there is an interesting title, the demand rises up to 200-250 pieces (once per 10 days in average).
This event can not be predicted! Order is made every day. As per our agreement with my dealer If I can not sell any title I can return 30 pieces of it and I do not need to pay CZK 20,- for these DVDs.

THE GOAL OF THE SIMULATION

The main objective of this simulation is to find the optimum number DVD order in whichthe average gain will be maximized.

THE TYPE OF SIMULATION / TOOL / METHOD

Monte Carlo simulation method / MS Excel 2013.
: Who has written this? Sounds too simple to me for a seminal paper. [[User:Tomáš|Tomáš]] ([[User talk:Tomáš|talk]]) 13:23, 2 January 2016 (CET)

Load-balancing

2016-01-17T22:49:56Z

Xtomp36: /* Results */

*'''Project name:''' Load-balancing
*'''Class:''' 4IT496 (WS 2015/2016)
*'''Author:''' Bc. Patrik Tomášek
*'''Model type:''' Discrete-event simulation
*'''Software used:''' SimProcess, trial version

=Problem definition=

A hosting company with its own infrastructure is using so called "load balancing" to distribute the overall load between multiple servers (hardware nodes) and “high-availability” to minimize service down-time.

'''Load-balancing''' 
In the case of web hosting service it means dividing the incoming request between multiple devices (called nodes) based on a set of rules (priority, weight, etc). The simulation is based on nginx (a webserver software) load balancing.
 
[[File:Nginx_load_balancing.png]]
 
''Note: In the picture the load-balacer isn't redundant, therefore HA isn't enabled. The simulation has two load-balances present.''


'''High-avaibility''' 
Ensures that a system or component is operational for desirable time. The solution necessary to provide web hosting consists of many parts, where all of them need to be on-line for the whole to be operational.

To enable HA a provider can use failover and backups. Failover is basically a backup piece of hardware, which ensures that when a component goes off-line another takes it's place. After that it's necessary to load the backup on the component that took over. If there is a SAN (storage area network) implemented than there is no need to load a backup, because failover just uses the same data from one central storage, which is used for all the server nodes. In the simulation a SAN is implemented and failover is taken into consideration.


'''Anti-DoS/DDoS''' 
Denial-of-service (DoS) attack is an incident is witch the targeted service goes down. Distributed denial-of-service means, that more than one system is used to attack a single target. There are more means of possible protection against such attack. Setting up a decent firewall rules might be a good place to start, but it isn't so effective as implementing a device, which can mitigate the attack. A Radware defencePro device is implemented in the simulation.



The goal of the simulation is to find the optimal number of server and other components necessary to enable the mentioned functions (LB, HA, Anti-DoS).


=Method=
Discrete event simulation can be solved using other software than SimProcess, however graphical interface was preferred. It is more user friendly to use GUI to create the simulation than typing it in code. 
There are some limitations present due to use of a trial revision, but this simulation doesn't reached them. 
All the prices in the simulation are in the default currency, which is USD. 
At first I wanted to due a simulation of whole day, however the default time unit is a second and the simulation took a lot of time. So there are three small time intervals simulated with the highest number of requests needed to be processed taken into consideration.

=Model=


The simulation is divided into 4 main processes: 
'''- New requests''' - generation of new requests, further information in Entities section. 
'''- DoS mitigation''' - implementation of anti-dos device. 
'''- Load balancing''' - main process, witch distributes the generated (incoming) request between resources. 
'''- Served request''' - simple dispose of the generated requests. 
[[File:Whole_model.jpg]]



The simulation is set to run in two 9 minutes iterations, using multiple schedules to apply the content changes of a web page (cached/un-cached content). 
Generated request are the highest recorded numbers of the given hosting environment (so called peak). Other possibilities doesn't need to be taken into consideration because the system must be able to handle the peak and with owned hardware, there is not much use for downscaling. Besides the peak might come at odd hour, and starting up an off-line server takes a lot of time. It would be better to use the unallocated server usage for some other calculation, which might be beneficial for the provider. 
'''Time unit used''': seconds 
'''Number of all requests:''' 9000 per second is the maximum recorded in the given environment. The number was divided by 100 for the simulation purposes (so it's 90). 
'''Number of replications:''' 2

== Entities ==
This is a list of all defined and used entities within the simulation. 


'''Requests''' 
The request data are based on real data from an hosting environment, that hosts multiple Magento e-shops (a rather complex system, which uses a lot of hardware). 
The incoming request are categorised because Magento uses caching and indexing. It takes less time to server a cached content than un-cached one. Further more there are multiple schedules used to generate the requests. If new data have been added to the e-shop, that they need to be cached first, witch means more large requests.


'''- Small request''' 
Request of a cached content.

'''- Standard request''' 
Request of a content on storage server, no need to accesses database.

'''- Large request''' 
Request of a content on storage server and database.

'''D-DoS''' 
A large number of requests aiming to cripple the system.


== Resources ==

''This is only a brief summary of resources used, further information follows in processes section of this page.'' 
When it comes to resource price, I take only lease into consideration. Of course it is also possible to purchase the given component. Cloud computing is also not considered.

'''Load-balancer''' 
An important component described earlier on this page. 
These devices are quite expensive. In case of lease, the price can be 300 USD monthly for a sufficient component for this simulation requirements.

'''Web server''' 
There are two types of this resource, each with different capabilities: 
Type 1 - a faster server that goes for 150 USD monthly 
Type 2 - a slower server that goes for 100 USD monthly 

'''Storage server SAN''' 
In the simulation only SAN is used as a name of this resource.
The price of one SAN depends on the configuration (number of HDDs, etc.). A sufficient SAN can be leased for 220 USD monthly.

'''Database server''' 
Database server is quite fast, so in the simulation only one is necessary, however to achieve HA even with DB server a second one should be implemented.
The price of one DB server is about 200 USD monthly.

'''Radware DefensePro''' 
This is an anti-dos component implemented in the simulation. 
The cost of this hardware is extreme, therefore partial lease is used. 
The price depends on the internet connection of the provider, is this case it goes around 50 USD monthly. 


== Processes ==

'''New requests''' 

[[File:Request_generate.jpg]]

In this process new requests are generated according to the following schedule: 
'''Interval:''' 1 second 

{| class="wikitable"
|-
! Entity type
! Most content cached
! Standard content
! New content added
! Sum of normal requests
! Addition DoS requests (red square)
|-
| New small request (green dot)
| Poi(45)
| Poi(25)
| Poi(13.5)
| Poi(90)
| Poi(150)
|-
| New standard request (orange dot)
| Poi(31.5)
| Poi(45)
| Poi(31.5)
| Poi(90)
| Poi(150)
|-
| New large request (black dot)
| Poi(13.5)
| Poi(20)
| Poi(45)
| Poi(90)
| Poi(150)
|-
|}
 


'''DoS mittigation''' 

[[File:Dos_mittigation.jpg]]

This is a simple process, which implements anti-dos device and ensures that the system is not effected by the attack. 
The "Branch" activity simply lets forward only the correct request and the rest is disposed, thus the attack is mitigated. 
Without this device implemented the system would get flooded as later illustrated in results section of this page.


'''Load balancing''' 

[[File:Load balancing.jpg]]


This is the main process of the simulation. Firstly it is expected that about 1 % of incoming request are lost somewhere on the way, so that is what the "Route" activity does. The lost requests are simply disposed. Than the request arrives at Load-balancer, which divides them based on probability. It would be also possible to use least processing or something similar. The load-balancers are designed to handle extreme amount of requests and are very fast, therefore there is only as little delay as 1 ms in the simulation and event that is probably to high. Basically with the number of generated request in the simulation the LB can't be overloaded even in case of DoS attack. However it is a critical piece of hardware, because if it goes down, no request will reach the host, therefore it must be redundant to enable HA (so two LBs are required). 
After this the request reach either Server of type 1 or type 2. Server type one is a delay of Nor(25,3) ms and server type 2 is a delay of Nor(40,3.0) ms, because it is slower. Upon the web server processes the request it is decided where the request goes next based on Entity.type. In case of a small request it goes straight to the next Branch "Response". The standard and large request are send to storage server. The storage server represents another delay an because the system uses SAN it is a common resource for all the web servers. The delay of SAN is defined as Nor(10.0,2.0) ms. Standard request is than send to "Response" activity by "Entity route 2" activity. Large requests are further more send to DB server (the same delay as in the case of storage server) by the same activity. After that they continue to "Response" activity. Branch activity "Response" serves the content to the client, but only 90 % of requests are correct, 7 % are incorrect and client gives up and 3 % are also incorrect but client send a new request.
 


'''Requests served''' 
[[File:Served_requests.jpg]]

A simple dispose of served requests.


=Results=
This section is divided into the following cases.

<h2>Case 1: Optimal number of servers with DoS attack mitigation present</h2>

The simulation has shown that the optimal number of server necessary to process incoming request properly is 3 of type one and 2 of type two.

[[File:Lb_result_1.jpg]] 
Shows the number of all generated and processed requests.


[[File:Lb_result_2.jpg]] 
The wait time for resource is within the acceptable bounds. If it would be over 100 ms, some improvement should be done.


[[File:Lb_result_3.jpg]] 
The HA must be considered, but in this case it is expected that no more than one server will do off-line in one time. The probability of more than one server going off-line at the same time is quite low. the number of servers could be raised to achieve better HA capabilities, but it wont speed up the process much and there is a reserve in current state.


The cost of this system configuration is: 
Servers type one: 3x 150 USD 
Servers type two: 2x 100 USD 
Load-balancers: 2x 300 USD 
SANs: 2x 220 USD 
DB server: 1x 200 USD 
Total monthly costs: 1890 USD 
The entire solution is very expensive and it might be more cost effective to buy at least some of the devices. Web servers are relatively cheap, so I would start there.

<h2>Case 2: DoS attack with no mitigation</h2>
In this case the number of used resources is the same as in previous one, but the DoS mitigation devices is not present.

[[File:Lb_result_2_1.jpg]] 
Shows the number of all generated and processed requests. In here it is quite visible that the system was flooded and wasn't able to process all the requests.


[[File:Lb_result_2_2.jpg]] 
The wait time for resource is absolutely unacceptable, basically all the requests would time out before they would even start being processed.


[[File:Lb_result_2_3.jpg]] 
All the web servers are flooded and cant handle the number of incoming requests.


This shows that a DoS attack cant crush the entire system so proper measures must be in place. As is illustrated in previous case.

<h2>Case 3: same as case 1 with one server down (type 1)</h2>

[[File:Lb_result_3_1.jpg]] 
Some request were not processed, due to one server being off-line, but they would if the simulation would continue for a few more milliseconds. 
This shows that even with one server down all the requests are processed and HA is functional.


[[File:Lb_result_3_2.jpg]] 
One off-line server raised the "Type_one_servers" resource usage from 72 % to 93 %, which is still sufficient.



=Conclusion=

The simulation has identified the optimal number of web servers and other involved components. As expected the costs of such system are quite high. 
Because the simulation was made using SimProcess, there are some limitations as to defining the resource limitations. In here it represents only the delays and number of units available. It would be better to represent the server capabilities by something else than a delay, however it is sufficient. 
The simulation can be modified to find optimal number of servers and other components for a different number of requests than used. In this simulation the optimal numbers are: 3 web servers of type one, 2 web servery of type two, two load-balancers (only due to redundancy, otherwise one would be enough), two SANs (same reason), one DB. The overall monthly cost of such a system is about 1890 USD. 
The simulation has also illustrated the effect of a DoS attack and a possible protection against it. The additional costs for such a protection depends on too many factors. However in some datacenters it is possible to lease a part of the device, that does the mitigation (it could start at 50 USD per month).

=Code=
[[File:Load_balancing.spm]]

=Sources=
http://searchnetworking.techtarget.com/definition/load-balancing 
http://searchstorage.techtarget.com/definition/failover 
http://searchsoftwarequality.techtarget.com/definition/denial-of-service 
http://searchsecurity.techtarget.com/definition/distributed-denial-of-service-attack 
http://cepa.io/devlog/secure-https-load-balancing-with-nginx

File:Lb result 3 2.jpg

2016-01-17T22:46:33Z

Xtomp36:

File:Lb result 3 1.jpg

2016-01-17T22:46:22Z

Xtomp36:

Load-balancing

2016-01-17T22:38:25Z

Xtomp36: /* Method */

*'''Project name:''' Load-balancing
*'''Class:''' 4IT496 (WS 2015/2016)
*'''Author:''' Bc. Patrik Tomášek
*'''Model type:''' Discrete-event simulation
*'''Software used:''' SimProcess, trial version

=Problem definition=

A hosting company with its own infrastructure is using so called "load balancing" to distribute the overall load between multiple servers (hardware nodes) and “high-availability” to minimize service down-time.

'''Load-balancing''' 
In the case of web hosting service it means dividing the incoming request between multiple devices (called nodes) based on a set of rules (priority, weight, etc). The simulation is based on nginx (a webserver software) load balancing.
 
[[File:Nginx_load_balancing.png]]
 
''Note: In the picture the load-balacer isn't redundant, therefore HA isn't enabled. The simulation has two load-balances present.''


'''High-avaibility''' 
Ensures that a system or component is operational for desirable time. The solution necessary to provide web hosting consists of many parts, where all of them need to be on-line for the whole to be operational.

To enable HA a provider can use failover and backups. Failover is basically a backup piece of hardware, which ensures that when a component goes off-line another takes it's place. After that it's necessary to load the backup on the component that took over. If there is a SAN (storage area network) implemented than there is no need to load a backup, because failover just uses the same data from one central storage, which is used for all the server nodes. In the simulation a SAN is implemented and failover is taken into consideration.


'''Anti-DoS/DDoS''' 
Denial-of-service (DoS) attack is an incident is witch the targeted service goes down. Distributed denial-of-service means, that more than one system is used to attack a single target. There are more means of possible protection against such attack. Setting up a decent firewall rules might be a good place to start, but it isn't so effective as implementing a device, which can mitigate the attack. A Radware defencePro device is implemented in the simulation.



The goal of the simulation is to find the optimal number of server and other components necessary to enable the mentioned functions (LB, HA, Anti-DoS).


=Method=
Discrete event simulation can be solved using other software than SimProcess, however graphical interface was preferred. It is more user friendly to use GUI to create the simulation than typing it in code. 
There are some limitations present due to use of a trial revision, but this simulation doesn't reached them. 
All the prices in the simulation are in the default currency, which is USD. 
At first I wanted to due a simulation of whole day, however the default time unit is a second and the simulation took a lot of time. So there are three small time intervals simulated with the highest number of requests needed to be processed taken into consideration.

=Model=


The simulation is divided into 4 main processes: 
'''- New requests''' - generation of new requests, further information in Entities section. 
'''- DoS mitigation''' - implementation of anti-dos device. 
'''- Load balancing''' - main process, witch distributes the generated (incoming) request between resources. 
'''- Served request''' - simple dispose of the generated requests. 
[[File:Whole_model.jpg]]



The simulation is set to run in two 9 minutes iterations, using multiple schedules to apply the content changes of a web page (cached/un-cached content). 
Generated request are the highest recorded numbers of the given hosting environment (so called peak). Other possibilities doesn't need to be taken into consideration because the system must be able to handle the peak and with owned hardware, there is not much use for downscaling. Besides the peak might come at odd hour, and starting up an off-line server takes a lot of time. It would be better to use the unallocated server usage for some other calculation, which might be beneficial for the provider. 
'''Time unit used''': seconds 
'''Number of all requests:''' 9000 per second is the maximum recorded in the given environment. The number was divided by 100 for the simulation purposes (so it's 90). 
'''Number of replications:''' 2

== Entities ==
This is a list of all defined and used entities within the simulation. 


'''Requests''' 
The request data are based on real data from an hosting environment, that hosts multiple Magento e-shops (a rather complex system, which uses a lot of hardware). 
The incoming request are categorised because Magento uses caching and indexing. It takes less time to server a cached content than un-cached one. Further more there are multiple schedules used to generate the requests. If new data have been added to the e-shop, that they need to be cached first, witch means more large requests.


'''- Small request''' 
Request of a cached content.

'''- Standard request''' 
Request of a content on storage server, no need to accesses database.

'''- Large request''' 
Request of a content on storage server and database.

'''D-DoS''' 
A large number of requests aiming to cripple the system.


== Resources ==

''This is only a brief summary of resources used, further information follows in processes section of this page.'' 
When it comes to resource price, I take only lease into consideration. Of course it is also possible to purchase the given component. Cloud computing is also not considered.

'''Load-balancer''' 
An important component described earlier on this page. 
These devices are quite expensive. In case of lease, the price can be 300 USD monthly for a sufficient component for this simulation requirements.

'''Web server''' 
There are two types of this resource, each with different capabilities: 
Type 1 - a faster server that goes for 150 USD monthly 
Type 2 - a slower server that goes for 100 USD monthly 

'''Storage server SAN''' 
In the simulation only SAN is used as a name of this resource.
The price of one SAN depends on the configuration (number of HDDs, etc.). A sufficient SAN can be leased for 220 USD monthly.

'''Database server''' 
Database server is quite fast, so in the simulation only one is necessary, however to achieve HA even with DB server a second one should be implemented.
The price of one DB server is about 200 USD monthly.

'''Radware DefensePro''' 
This is an anti-dos component implemented in the simulation. 
The cost of this hardware is extreme, therefore partial lease is used. 
The price depends on the internet connection of the provider, is this case it goes around 50 USD monthly. 


== Processes ==

'''New requests''' 

[[File:Request_generate.jpg]]

In this process new requests are generated according to the following schedule: 
'''Interval:''' 1 second 

{| class="wikitable"
|-
! Entity type
! Most content cached
! Standard content
! New content added
! Sum of normal requests
! Addition DoS requests (red square)
|-
| New small request (green dot)
| Poi(45)
| Poi(25)
| Poi(13.5)
| Poi(90)
| Poi(150)
|-
| New standard request (orange dot)
| Poi(31.5)
| Poi(45)
| Poi(31.5)
| Poi(90)
| Poi(150)
|-
| New large request (black dot)
| Poi(13.5)
| Poi(20)
| Poi(45)
| Poi(90)
| Poi(150)
|-
|}
 


'''DoS mittigation''' 

[[File:Dos_mittigation.jpg]]

This is a simple process, which implements anti-dos device and ensures that the system is not effected by the attack. 
The "Branch" activity simply lets forward only the correct request and the rest is disposed, thus the attack is mitigated. 
Without this device implemented the system would get flooded as later illustrated in results section of this page.


'''Load balancing''' 

[[File:Load balancing.jpg]]


This is the main process of the simulation. Firstly it is expected that about 1 % of incoming request are lost somewhere on the way, so that is what the "Route" activity does. The lost requests are simply disposed. Than the request arrives at Load-balancer, which divides them based on probability. It would be also possible to use least processing or something similar. The load-balancers are designed to handle extreme amount of requests and are very fast, therefore there is only as little delay as 1 ms in the simulation and event that is probably to high. Basically with the number of generated request in the simulation the LB can't be overloaded even in case of DoS attack. However it is a critical piece of hardware, because if it goes down, no request will reach the host, therefore it must be redundant to enable HA (so two LBs are required). 
After this the request reach either Server of type 1 or type 2. Server type one is a delay of Nor(25,3) ms and server type 2 is a delay of Nor(40,3.0) ms, because it is slower. Upon the web server processes the request it is decided where the request goes next based on Entity.type. In case of a small request it goes straight to the next Branch "Response". The standard and large request are send to storage server. The storage server represents another delay an because the system uses SAN it is a common resource for all the web servers. The delay of SAN is defined as Nor(10.0,2.0) ms. Standard request is than send to "Response" activity by "Entity route 2" activity. Large requests are further more send to DB server (the same delay as in the case of storage server) by the same activity. After that they continue to "Response" activity. Branch activity "Response" serves the content to the client, but only 90 % of requests are correct, 7 % are incorrect and client gives up and 3 % are also incorrect but client send a new request.
 


'''Requests served''' 
[[File:Served_requests.jpg]]

A simple dispose of served requests.


=Results=
This section is divided into the following cases.

<h2>Optimal number of servers with DoS attack mitigation present</h2>

The simulation has shown that the optimal number of server necessary to process incoming request properly is 3 of type one and 2 of type two.

[[File:Lb_result_1.jpg]] 
Shows the number of all generated and processed requests.


[[File:Lb_result_2.jpg]] 
The wait time for resource is within the acceptable bounds. If it would be over 100 ms, some improvement should be done.


[[File:Lb_result_3.jpg]] 
The HA must be considered, but in this case it is expected that no more than one server will do off-line in one time. The probability of more than one server going off-line at the same time is quite low. the number of servers could be raised to achieve better HA capabilities, but it wont speed up the process much and there is a reserve in current state.


The cost of this system configuration is: 
Servers type one: 3x 150 USD 
Servers type two: 2x 100 USD 
Load-balancers: 2x 300 USD 
SANs: 2x 220 USD 
DB server: 1x 200 USD 
Total monthly costs: 1890 USD 
The entire solution is very expensive and it might be more cost effective to buy at least some of the devices. Web servers are relatively cheap, so I would start there.

<h2>Case of DoS attack with no mitigation</h2>
In this case the number of used resources is the same as in previous one, but the DoS mitigation devices is not present.

[[File:Lb_result_2_1.jpg]] 
Shows the number of all generated and processed requests. In here it is quite visible that the system was flooded and wasn't able to process all the requests.


[[File:Lb_result_2_2.jpg]] 
The wait time for resource is absolutely unacceptable, basically all the requests would time out before they would even start being processed.


[[File:Lb_result_2_3.jpg]] 
All the web servers are flooded and cant handle the number of incoming requests.


This shows that a DoS attack cant crush the entire system so proper measures must be in place. As is illustrated in previous case.

=Conclusion=

The simulation has identified the optimal number of web servers and other involved components. As expected the costs of such system are quite high. 
Because the simulation was made using SimProcess, there are some limitations as to defining the resource limitations. In here it represents only the delays and number of units available. It would be better to represent the server capabilities by something else than a delay, however it is sufficient. 
The simulation can be modified to find optimal number of servers and other components for a different number of requests than used. In this simulation the optimal numbers are: 3 web servers of type one, 2 web servery of type two, two load-balancers (only due to redundancy, otherwise one would be enough), two SANs (same reason), one DB. The overall monthly cost of such a system is about 1890 USD. 
The simulation has also illustrated the effect of a DoS attack and a possible protection against it. The additional costs for such a protection depends on too many factors. However in some datacenters it is possible to lease a part of the device, that does the mitigation (it could start at 50 USD per month).

=Code=
[[File:Load_balancing.spm]]

=Sources=
http://searchnetworking.techtarget.com/definition/load-balancing 
http://searchstorage.techtarget.com/definition/failover 
http://searchsoftwarequality.techtarget.com/definition/denial-of-service 
http://searchsecurity.techtarget.com/definition/distributed-denial-of-service-attack 
http://cepa.io/devlog/secure-https-load-balancing-with-nginx

Load-balancing

2016-01-17T22:36:37Z

Xtomp36: /* Method */

*'''Project name:''' Load-balancing
*'''Class:''' 4IT496 (WS 2015/2016)
*'''Author:''' Bc. Patrik Tomášek
*'''Model type:''' Discrete-event simulation
*'''Software used:''' SimProcess, trial version

=Problem definition=

A hosting company with its own infrastructure is using so called "load balancing" to distribute the overall load between multiple servers (hardware nodes) and “high-availability” to minimize service down-time.

'''Load-balancing''' 
In the case of web hosting service it means dividing the incoming request between multiple devices (called nodes) based on a set of rules (priority, weight, etc). The simulation is based on nginx (a webserver software) load balancing.
 
[[File:Nginx_load_balancing.png]]
 
''Note: In the picture the load-balacer isn't redundant, therefore HA isn't enabled. The simulation has two load-balances present.''


'''High-avaibility''' 
Ensures that a system or component is operational for desirable time. The solution necessary to provide web hosting consists of many parts, where all of them need to be on-line for the whole to be operational.

To enable HA a provider can use failover and backups. Failover is basically a backup piece of hardware, which ensures that when a component goes off-line another takes it's place. After that it's necessary to load the backup on the component that took over. If there is a SAN (storage area network) implemented than there is no need to load a backup, because failover just uses the same data from one central storage, which is used for all the server nodes. In the simulation a SAN is implemented and failover is taken into consideration.


'''Anti-DoS/DDoS''' 
Denial-of-service (DoS) attack is an incident is witch the targeted service goes down. Distributed denial-of-service means, that more than one system is used to attack a single target. There are more means of possible protection against such attack. Setting up a decent firewall rules might be a good place to start, but it isn't so effective as implementing a device, which can mitigate the attack. A Radware defencePro device is implemented in the simulation.



The goal of the simulation is to find the optimal number of server and other components necessary to enable the mentioned functions (LB, HA, Anti-DoS).


=Method=
Discrete event simulation can be solved using other software than SimProcess, however graphical interface was preferred. It is more user friendly to use GUI to create the simulation than typing it in code. 
There are some limitations present due to use of a trial revision, but this simulation doesn't reached them. 
All the prices in the simulation are in the default currency, which is USD. 
At first I wanted to due a simulation of whole day, however the default time unit is in seconds and the simulation took a lot of time. Therefore there are only tree small time intervals simulated with the highest number of requests needed to be processed taken into consideration.

=Model=


The simulation is divided into 4 main processes: 
'''- New requests''' - generation of new requests, further information in Entities section. 
'''- DoS mitigation''' - implementation of anti-dos device. 
'''- Load balancing''' - main process, witch distributes the generated (incoming) request between resources. 
'''- Served request''' - simple dispose of the generated requests. 
[[File:Whole_model.jpg]]



The simulation is set to run in two 9 minutes iterations, using multiple schedules to apply the content changes of a web page (cached/un-cached content). 
Generated request are the highest recorded numbers of the given hosting environment (so called peak). Other possibilities doesn't need to be taken into consideration because the system must be able to handle the peak and with owned hardware, there is not much use for downscaling. Besides the peak might come at odd hour, and starting up an off-line server takes a lot of time. It would be better to use the unallocated server usage for some other calculation, which might be beneficial for the provider. 
'''Time unit used''': seconds 
'''Number of all requests:''' 9000 per second is the maximum recorded in the given environment. The number was divided by 100 for the simulation purposes (so it's 90). 
'''Number of replications:''' 2

== Entities ==
This is a list of all defined and used entities within the simulation. 


'''Requests''' 
The request data are based on real data from an hosting environment, that hosts multiple Magento e-shops (a rather complex system, which uses a lot of hardware). 
The incoming request are categorised because Magento uses caching and indexing. It takes less time to server a cached content than un-cached one. Further more there are multiple schedules used to generate the requests. If new data have been added to the e-shop, that they need to be cached first, witch means more large requests.


'''- Small request''' 
Request of a cached content.

'''- Standard request''' 
Request of a content on storage server, no need to accesses database.

'''- Large request''' 
Request of a content on storage server and database.

'''D-DoS''' 
A large number of requests aiming to cripple the system.


== Resources ==

''This is only a brief summary of resources used, further information follows in processes section of this page.'' 
When it comes to resource price, I take only lease into consideration. Of course it is also possible to purchase the given component. Cloud computing is also not considered.

'''Load-balancer''' 
An important component described earlier on this page. 
These devices are quite expensive. In case of lease, the price can be 300 USD monthly for a sufficient component for this simulation requirements.

'''Web server''' 
There are two types of this resource, each with different capabilities: 
Type 1 - a faster server that goes for 150 USD monthly 
Type 2 - a slower server that goes for 100 USD monthly 

'''Storage server SAN''' 
In the simulation only SAN is used as a name of this resource.
The price of one SAN depends on the configuration (number of HDDs, etc.). A sufficient SAN can be leased for 220 USD monthly.

'''Database server''' 
Database server is quite fast, so in the simulation only one is necessary, however to achieve HA even with DB server a second one should be implemented.
The price of one DB server is about 200 USD monthly.

'''Radware DefensePro''' 
This is an anti-dos component implemented in the simulation. 
The cost of this hardware is extreme, therefore partial lease is used. 
The price depends on the internet connection of the provider, is this case it goes around 50 USD monthly. 


== Processes ==

'''New requests''' 

[[File:Request_generate.jpg]]

In this process new requests are generated according to the following schedule: 
'''Interval:''' 1 second 

{| class="wikitable"
|-
! Entity type
! Most content cached
! Standard content
! New content added
! Sum of normal requests
! Addition DoS requests (red square)
|-
| New small request (green dot)
| Poi(45)
| Poi(25)
| Poi(13.5)
| Poi(90)
| Poi(150)
|-
| New standard request (orange dot)
| Poi(31.5)
| Poi(45)
| Poi(31.5)
| Poi(90)
| Poi(150)
|-
| New large request (black dot)
| Poi(13.5)
| Poi(20)
| Poi(45)
| Poi(90)
| Poi(150)
|-
|}
 


'''DoS mittigation''' 

[[File:Dos_mittigation.jpg]]

This is a simple process, which implements anti-dos device and ensures that the system is not effected by the attack. 
The "Branch" activity simply lets forward only the correct request and the rest is disposed, thus the attack is mitigated. 
Without this device implemented the system would get flooded as later illustrated in results section of this page.


'''Load balancing''' 

[[File:Load balancing.jpg]]


This is the main process of the simulation. Firstly it is expected that about 1 % of incoming request are lost somewhere on the way, so that is what the "Route" activity does. The lost requests are simply disposed. Than the request arrives at Load-balancer, which divides them based on probability. It would be also possible to use least processing or something similar. The load-balancers are designed to handle extreme amount of requests and are very fast, therefore there is only as little delay as 1 ms in the simulation and event that is probably to high. Basically with the number of generated request in the simulation the LB can't be overloaded even in case of DoS attack. However it is a critical piece of hardware, because if it goes down, no request will reach the host, therefore it must be redundant to enable HA (so two LBs are required). 
After this the request reach either Server of type 1 or type 2. Server type one is a delay of Nor(25,3) ms and server type 2 is a delay of Nor(40,3.0) ms, because it is slower. Upon the web server processes the request it is decided where the request goes next based on Entity.type. In case of a small request it goes straight to the next Branch "Response". The standard and large request are send to storage server. The storage server represents another delay an because the system uses SAN it is a common resource for all the web servers. The delay of SAN is defined as Nor(10.0,2.0) ms. Standard request is than send to "Response" activity by "Entity route 2" activity. Large requests are further more send to DB server (the same delay as in the case of storage server) by the same activity. After that they continue to "Response" activity. Branch activity "Response" serves the content to the client, but only 90 % of requests are correct, 7 % are incorrect and client gives up and 3 % are also incorrect but client send a new request.
 


'''Requests served''' 
[[File:Served_requests.jpg]]

A simple dispose of served requests.


=Results=
This section is divided into the following cases.

<h2>Optimal number of servers with DoS attack mitigation present</h2>

The simulation has shown that the optimal number of server necessary to process incoming request properly is 3 of type one and 2 of type two.

[[File:Lb_result_1.jpg]] 
Shows the number of all generated and processed requests.


[[File:Lb_result_2.jpg]] 
The wait time for resource is within the acceptable bounds. If it would be over 100 ms, some improvement should be done.


[[File:Lb_result_3.jpg]] 
The HA must be considered, but in this case it is expected that no more than one server will do off-line in one time. The probability of more than one server going off-line at the same time is quite low. the number of servers could be raised to achieve better HA capabilities, but it wont speed up the process much and there is a reserve in current state.


The cost of this system configuration is: 
Servers type one: 3x 150 USD 
Servers type two: 2x 100 USD 
Load-balancers: 2x 300 USD 
SANs: 2x 220 USD 
DB server: 1x 200 USD 
Total monthly costs: 1890 USD 
The entire solution is very expensive and it might be more cost effective to buy at least some of the devices. Web servers are relatively cheap, so I would start there.

<h2>Case of DoS attack with no mitigation</h2>
In this case the number of used resources is the same as in previous one, but the DoS mitigation devices is not present.

[[File:Lb_result_2_1.jpg]] 
Shows the number of all generated and processed requests. In here it is quite visible that the system was flooded and wasn't able to process all the requests.


[[File:Lb_result_2_2.jpg]] 
The wait time for resource is absolutely unacceptable, basically all the requests would time out before they would even start being processed.


[[File:Lb_result_2_3.jpg]] 
All the web servers are flooded and cant handle the number of incoming requests.


This shows that a DoS attack cant crush the entire system so proper measures must be in place. As is illustrated in previous case.

=Conclusion=

The simulation has identified the optimal number of web servers and other involved components. As expected the costs of such system are quite high. 
Because the simulation was made using SimProcess, there are some limitations as to defining the resource limitations. In here it represents only the delays and number of units available. It would be better to represent the server capabilities by something else than a delay, however it is sufficient. 
The simulation can be modified to find optimal number of servers and other components for a different number of requests than used. In this simulation the optimal numbers are: 3 web servers of type one, 2 web servery of type two, two load-balancers (only due to redundancy, otherwise one would be enough), two SANs (same reason), one DB. The overall monthly cost of such a system is about 1890 USD. 
The simulation has also illustrated the effect of a DoS attack and a possible protection against it. The additional costs for such a protection depends on too many factors. However in some datacenters it is possible to lease a part of the device, that does the mitigation (it could start at 50 USD per month).

=Code=
[[File:Load_balancing.spm]]

=Sources=
http://searchnetworking.techtarget.com/definition/load-balancing 
http://searchstorage.techtarget.com/definition/failover 
http://searchsoftwarequality.techtarget.com/definition/denial-of-service 
http://searchsecurity.techtarget.com/definition/distributed-denial-of-service-attack 
http://cepa.io/devlog/secure-https-load-balancing-with-nginx

Load-balancing

2016-01-17T22:34:55Z

Xtomp36: /* Method */

*'''Project name:''' Load-balancing
*'''Class:''' 4IT496 (WS 2015/2016)
*'''Author:''' Bc. Patrik Tomášek
*'''Model type:''' Discrete-event simulation
*'''Software used:''' SimProcess, trial version

=Problem definition=

A hosting company with its own infrastructure is using so called "load balancing" to distribute the overall load between multiple servers (hardware nodes) and “high-availability” to minimize service down-time.

'''Load-balancing''' 
In the case of web hosting service it means dividing the incoming request between multiple devices (called nodes) based on a set of rules (priority, weight, etc). The simulation is based on nginx (a webserver software) load balancing.
 
[[File:Nginx_load_balancing.png]]
 
''Note: In the picture the load-balacer isn't redundant, therefore HA isn't enabled. The simulation has two load-balances present.''


'''High-avaibility''' 
Ensures that a system or component is operational for desirable time. The solution necessary to provide web hosting consists of many parts, where all of them need to be on-line for the whole to be operational.

To enable HA a provider can use failover and backups. Failover is basically a backup piece of hardware, which ensures that when a component goes off-line another takes it's place. After that it's necessary to load the backup on the component that took over. If there is a SAN (storage area network) implemented than there is no need to load a backup, because failover just uses the same data from one central storage, which is used for all the server nodes. In the simulation a SAN is implemented and failover is taken into consideration.


'''Anti-DoS/DDoS''' 
Denial-of-service (DoS) attack is an incident is witch the targeted service goes down. Distributed denial-of-service means, that more than one system is used to attack a single target. There are more means of possible protection against such attack. Setting up a decent firewall rules might be a good place to start, but it isn't so effective as implementing a device, which can mitigate the attack. A Radware defencePro device is implemented in the simulation.



The goal of the simulation is to find the optimal number of server and other components necessary to enable the mentioned functions (LB, HA, Anti-DoS).


=Method=
Discrete event simulation can be solved using other software than SimProcess, however graphical interface was preferred. It is more user friendly to use GUI to create the simulation than typing it in code. 
There are some limitations present due to use of a trial revision, but this simulation doesn't reached them. 
All the prices in the simulation are in the default currency, which is USD. 
At first I wanted to due a simulation of whole day, however the default time unit is in seconds and the simulation took a lot of time. Therefore only the highest number of requests needed to be process was taken into consideration and also some dividing was done.

=Model=


The simulation is divided into 4 main processes: 
'''- New requests''' - generation of new requests, further information in Entities section. 
'''- DoS mitigation''' - implementation of anti-dos device. 
'''- Load balancing''' - main process, witch distributes the generated (incoming) request between resources. 
'''- Served request''' - simple dispose of the generated requests. 
[[File:Whole_model.jpg]]



The simulation is set to run in two 9 minutes iterations, using multiple schedules to apply the content changes of a web page (cached/un-cached content). 
Generated request are the highest recorded numbers of the given hosting environment (so called peak). Other possibilities doesn't need to be taken into consideration because the system must be able to handle the peak and with owned hardware, there is not much use for downscaling. Besides the peak might come at odd hour, and starting up an off-line server takes a lot of time. It would be better to use the unallocated server usage for some other calculation, which might be beneficial for the provider. 
'''Time unit used''': seconds 
'''Number of all requests:''' 9000 per second is the maximum recorded in the given environment. The number was divided by 100 for the simulation purposes (so it's 90). 
'''Number of replications:''' 2

== Entities ==
This is a list of all defined and used entities within the simulation. 


'''Requests''' 
The request data are based on real data from an hosting environment, that hosts multiple Magento e-shops (a rather complex system, which uses a lot of hardware). 
The incoming request are categorised because Magento uses caching and indexing. It takes less time to server a cached content than un-cached one. Further more there are multiple schedules used to generate the requests. If new data have been added to the e-shop, that they need to be cached first, witch means more large requests.


'''- Small request''' 
Request of a cached content.

'''- Standard request''' 
Request of a content on storage server, no need to accesses database.

'''- Large request''' 
Request of a content on storage server and database.

'''D-DoS''' 
A large number of requests aiming to cripple the system.


== Resources ==

''This is only a brief summary of resources used, further information follows in processes section of this page.'' 
When it comes to resource price, I take only lease into consideration. Of course it is also possible to purchase the given component. Cloud computing is also not considered.

'''Load-balancer''' 
An important component described earlier on this page. 
These devices are quite expensive. In case of lease, the price can be 300 USD monthly for a sufficient component for this simulation requirements.

'''Web server''' 
There are two types of this resource, each with different capabilities: 
Type 1 - a faster server that goes for 150 USD monthly 
Type 2 - a slower server that goes for 100 USD monthly 

'''Storage server SAN''' 
In the simulation only SAN is used as a name of this resource.
The price of one SAN depends on the configuration (number of HDDs, etc.). A sufficient SAN can be leased for 220 USD monthly.

'''Database server''' 
Database server is quite fast, so in the simulation only one is necessary, however to achieve HA even with DB server a second one should be implemented.
The price of one DB server is about 200 USD monthly.

'''Radware DefensePro''' 
This is an anti-dos component implemented in the simulation. 
The cost of this hardware is extreme, therefore partial lease is used. 
The price depends on the internet connection of the provider, is this case it goes around 50 USD monthly. 


== Processes ==

'''New requests''' 

[[File:Request_generate.jpg]]

In this process new requests are generated according to the following schedule: 
'''Interval:''' 1 second 

{| class="wikitable"
|-
! Entity type
! Most content cached
! Standard content
! New content added
! Sum of normal requests
! Addition DoS requests (red square)
|-
| New small request (green dot)
| Poi(45)
| Poi(25)
| Poi(13.5)
| Poi(90)
| Poi(150)
|-
| New standard request (orange dot)
| Poi(31.5)
| Poi(45)
| Poi(31.5)
| Poi(90)
| Poi(150)
|-
| New large request (black dot)
| Poi(13.5)
| Poi(20)
| Poi(45)
| Poi(90)
| Poi(150)
|-
|}
 


'''DoS mittigation''' 

[[File:Dos_mittigation.jpg]]

This is a simple process, which implements anti-dos device and ensures that the system is not effected by the attack. 
The "Branch" activity simply lets forward only the correct request and the rest is disposed, thus the attack is mitigated. 
Without this device implemented the system would get flooded as later illustrated in results section of this page.


'''Load balancing''' 

[[File:Load balancing.jpg]]


This is the main process of the simulation. Firstly it is expected that about 1 % of incoming request are lost somewhere on the way, so that is what the "Route" activity does. The lost requests are simply disposed. Than the request arrives at Load-balancer, which divides them based on probability. It would be also possible to use least processing or something similar. The load-balancers are designed to handle extreme amount of requests and are very fast, therefore there is only as little delay as 1 ms in the simulation and event that is probably to high. Basically with the number of generated request in the simulation the LB can't be overloaded even in case of DoS attack. However it is a critical piece of hardware, because if it goes down, no request will reach the host, therefore it must be redundant to enable HA (so two LBs are required). 
After this the request reach either Server of type 1 or type 2. Server type one is a delay of Nor(25,3) ms and server type 2 is a delay of Nor(40,3.0) ms, because it is slower. Upon the web server processes the request it is decided where the request goes next based on Entity.type. In case of a small request it goes straight to the next Branch "Response". The standard and large request are send to storage server. The storage server represents another delay an because the system uses SAN it is a common resource for all the web servers. The delay of SAN is defined as Nor(10.0,2.0) ms. Standard request is than send to "Response" activity by "Entity route 2" activity. Large requests are further more send to DB server (the same delay as in the case of storage server) by the same activity. After that they continue to "Response" activity. Branch activity "Response" serves the content to the client, but only 90 % of requests are correct, 7 % are incorrect and client gives up and 3 % are also incorrect but client send a new request.
 


'''Requests served''' 
[[File:Served_requests.jpg]]

A simple dispose of served requests.


=Results=
This section is divided into the following cases.

<h2>Optimal number of servers with DoS attack mitigation present</h2>

The simulation has shown that the optimal number of server necessary to process incoming request properly is 3 of type one and 2 of type two.

[[File:Lb_result_1.jpg]] 
Shows the number of all generated and processed requests.


[[File:Lb_result_2.jpg]] 
The wait time for resource is within the acceptable bounds. If it would be over 100 ms, some improvement should be done.


[[File:Lb_result_3.jpg]] 
The HA must be considered, but in this case it is expected that no more than one server will do off-line in one time. The probability of more than one server going off-line at the same time is quite low. the number of servers could be raised to achieve better HA capabilities, but it wont speed up the process much and there is a reserve in current state.


The cost of this system configuration is: 
Servers type one: 3x 150 USD 
Servers type two: 2x 100 USD 
Load-balancers: 2x 300 USD 
SANs: 2x 220 USD 
DB server: 1x 200 USD 
Total monthly costs: 1890 USD 
The entire solution is very expensive and it might be more cost effective to buy at least some of the devices. Web servers are relatively cheap, so I would start there.

<h2>Case of DoS attack with no mitigation</h2>
In this case the number of used resources is the same as in previous one, but the DoS mitigation devices is not present.

[[File:Lb_result_2_1.jpg]] 
Shows the number of all generated and processed requests. In here it is quite visible that the system was flooded and wasn't able to process all the requests.


[[File:Lb_result_2_2.jpg]] 
The wait time for resource is absolutely unacceptable, basically all the requests would time out before they would even start being processed.


[[File:Lb_result_2_3.jpg]] 
All the web servers are flooded and cant handle the number of incoming requests.


This shows that a DoS attack cant crush the entire system so proper measures must be in place. As is illustrated in previous case.

=Conclusion=

The simulation has identified the optimal number of web servers and other involved components. As expected the costs of such system are quite high. 
Because the simulation was made using SimProcess, there are some limitations as to defining the resource limitations. In here it represents only the delays and number of units available. It would be better to represent the server capabilities by something else than a delay, however it is sufficient. 
The simulation can be modified to find optimal number of servers and other components for a different number of requests than used. In this simulation the optimal numbers are: 3 web servers of type one, 2 web servery of type two, two load-balancers (only due to redundancy, otherwise one would be enough), two SANs (same reason), one DB. The overall monthly cost of such a system is about 1890 USD. 
The simulation has also illustrated the effect of a DoS attack and a possible protection against it. The additional costs for such a protection depends on too many factors. However in some datacenters it is possible to lease a part of the device, that does the mitigation (it could start at 50 USD per month).

=Code=
[[File:Load_balancing.spm]]

=Sources=
http://searchnetworking.techtarget.com/definition/load-balancing 
http://searchstorage.techtarget.com/definition/failover 
http://searchsoftwarequality.techtarget.com/definition/denial-of-service 
http://searchsecurity.techtarget.com/definition/distributed-denial-of-service-attack 
http://cepa.io/devlog/secure-https-load-balancing-with-nginx

Load-balancing

2016-01-17T22:30:56Z

Xtomp36: /* Method */

*'''Project name:''' Load-balancing
*'''Class:''' 4IT496 (WS 2015/2016)
*'''Author:''' Bc. Patrik Tomášek
*'''Model type:''' Discrete-event simulation
*'''Software used:''' SimProcess, trial version

=Problem definition=

A hosting company with its own infrastructure is using so called "load balancing" to distribute the overall load between multiple servers (hardware nodes) and “high-availability” to minimize service down-time.

'''Load-balancing''' 
In the case of web hosting service it means dividing the incoming request between multiple devices (called nodes) based on a set of rules (priority, weight, etc). The simulation is based on nginx (a webserver software) load balancing.
 
[[File:Nginx_load_balancing.png]]
 
''Note: In the picture the load-balacer isn't redundant, therefore HA isn't enabled. The simulation has two load-balances present.''


'''High-avaibility''' 
Ensures that a system or component is operational for desirable time. The solution necessary to provide web hosting consists of many parts, where all of them need to be on-line for the whole to be operational.

To enable HA a provider can use failover and backups. Failover is basically a backup piece of hardware, which ensures that when a component goes off-line another takes it's place. After that it's necessary to load the backup on the component that took over. If there is a SAN (storage area network) implemented than there is no need to load a backup, because failover just uses the same data from one central storage, which is used for all the server nodes. In the simulation a SAN is implemented and failover is taken into consideration.


'''Anti-DoS/DDoS''' 
Denial-of-service (DoS) attack is an incident is witch the targeted service goes down. Distributed denial-of-service means, that more than one system is used to attack a single target. There are more means of possible protection against such attack. Setting up a decent firewall rules might be a good place to start, but it isn't so effective as implementing a device, which can mitigate the attack. A Radware defencePro device is implemented in the simulation.



The goal of the simulation is to find the optimal number of server and other components necessary to enable the mentioned functions (LB, HA, Anti-DoS).


=Method=
Discrete event simulation can be solved using other software than SimProcess, however graphical interface was preferred. It is more user friendly to use GUI to create the simulation than typing it in code. 
There are some limitations present due to use of a trial revision, but this simulation doesn't reached them. 
All the prices in the simulation are in the default currency, which is USD.

=Model=


The simulation is divided into 4 main processes: 
'''- New requests''' - generation of new requests, further information in Entities section. 
'''- DoS mitigation''' - implementation of anti-dos device. 
'''- Load balancing''' - main process, witch distributes the generated (incoming) request between resources. 
'''- Served request''' - simple dispose of the generated requests. 
[[File:Whole_model.jpg]]



The simulation is set to run in two 9 minutes iterations, using multiple schedules to apply the content changes of a web page (cached/un-cached content). 
Generated request are the highest recorded numbers of the given hosting environment (so called peak). Other possibilities doesn't need to be taken into consideration because the system must be able to handle the peak and with owned hardware, there is not much use for downscaling. Besides the peak might come at odd hour, and starting up an off-line server takes a lot of time. It would be better to use the unallocated server usage for some other calculation, which might be beneficial for the provider. 
'''Time unit used''': seconds 
'''Number of all requests:''' 9000 per second is the maximum recorded in the given environment. The number was divided by 100 for the simulation purposes (so it's 90). 
'''Number of replications:''' 2

== Entities ==
This is a list of all defined and used entities within the simulation. 


'''Requests''' 
The request data are based on real data from an hosting environment, that hosts multiple Magento e-shops (a rather complex system, which uses a lot of hardware). 
The incoming request are categorised because Magento uses caching and indexing. It takes less time to server a cached content than un-cached one. Further more there are multiple schedules used to generate the requests. If new data have been added to the e-shop, that they need to be cached first, witch means more large requests.


'''- Small request''' 
Request of a cached content.

'''- Standard request''' 
Request of a content on storage server, no need to accesses database.

'''- Large request''' 
Request of a content on storage server and database.

'''D-DoS''' 
A large number of requests aiming to cripple the system.


== Resources ==

''This is only a brief summary of resources used, further information follows in processes section of this page.'' 
When it comes to resource price, I take only lease into consideration. Of course it is also possible to purchase the given component. Cloud computing is also not considered.

'''Load-balancer''' 
An important component described earlier on this page. 
These devices are quite expensive. In case of lease, the price can be 300 USD monthly for a sufficient component for this simulation requirements.

'''Web server''' 
There are two types of this resource, each with different capabilities: 
Type 1 - a faster server that goes for 150 USD monthly 
Type 2 - a slower server that goes for 100 USD monthly 

'''Storage server SAN''' 
In the simulation only SAN is used as a name of this resource.
The price of one SAN depends on the configuration (number of HDDs, etc.). A sufficient SAN can be leased for 220 USD monthly.

'''Database server''' 
Database server is quite fast, so in the simulation only one is necessary, however to achieve HA even with DB server a second one should be implemented.
The price of one DB server is about 200 USD monthly.

'''Radware DefensePro''' 
This is an anti-dos component implemented in the simulation. 
The cost of this hardware is extreme, therefore partial lease is used. 
The price depends on the internet connection of the provider, is this case it goes around 50 USD monthly. 


== Processes ==

'''New requests''' 

[[File:Request_generate.jpg]]

In this process new requests are generated according to the following schedule: 
'''Interval:''' 1 second 

{| class="wikitable"
|-
! Entity type
! Most content cached
! Standard content
! New content added
! Sum of normal requests
! Addition DoS requests (red square)
|-
| New small request (green dot)
| Poi(45)
| Poi(25)
| Poi(13.5)
| Poi(90)
| Poi(150)
|-
| New standard request (orange dot)
| Poi(31.5)
| Poi(45)
| Poi(31.5)
| Poi(90)
| Poi(150)
|-
| New large request (black dot)
| Poi(13.5)
| Poi(20)
| Poi(45)
| Poi(90)
| Poi(150)
|-
|}
 


'''DoS mittigation''' 

[[File:Dos_mittigation.jpg]]

This is a simple process, which implements anti-dos device and ensures that the system is not effected by the attack. 
The "Branch" activity simply lets forward only the correct request and the rest is disposed, thus the attack is mitigated. 
Without this device implemented the system would get flooded as later illustrated in results section of this page.


'''Load balancing''' 

[[File:Load balancing.jpg]]


This is the main process of the simulation. Firstly it is expected that about 1 % of incoming request are lost somewhere on the way, so that is what the "Route" activity does. The lost requests are simply disposed. Than the request arrives at Load-balancer, which divides them based on probability. It would be also possible to use least processing or something similar. The load-balancers are designed to handle extreme amount of requests and are very fast, therefore there is only as little delay as 1 ms in the simulation and event that is probably to high. Basically with the number of generated request in the simulation the LB can't be overloaded even in case of DoS attack. However it is a critical piece of hardware, because if it goes down, no request will reach the host, therefore it must be redundant to enable HA (so two LBs are required). 
After this the request reach either Server of type 1 or type 2. Server type one is a delay of Nor(25,3) ms and server type 2 is a delay of Nor(40,3.0) ms, because it is slower. Upon the web server processes the request it is decided where the request goes next based on Entity.type. In case of a small request it goes straight to the next Branch "Response". The standard and large request are send to storage server. The storage server represents another delay an because the system uses SAN it is a common resource for all the web servers. The delay of SAN is defined as Nor(10.0,2.0) ms. Standard request is than send to "Response" activity by "Entity route 2" activity. Large requests are further more send to DB server (the same delay as in the case of storage server) by the same activity. After that they continue to "Response" activity. Branch activity "Response" serves the content to the client, but only 90 % of requests are correct, 7 % are incorrect and client gives up and 3 % are also incorrect but client send a new request.
 


'''Requests served''' 
[[File:Served_requests.jpg]]

A simple dispose of served requests.


=Results=
This section is divided into the following cases.

<h2>Optimal number of servers with DoS attack mitigation present</h2>

The simulation has shown that the optimal number of server necessary to process incoming request properly is 3 of type one and 2 of type two.

[[File:Lb_result_1.jpg]] 
Shows the number of all generated and processed requests.


[[File:Lb_result_2.jpg]] 
The wait time for resource is within the acceptable bounds. If it would be over 100 ms, some improvement should be done.


[[File:Lb_result_3.jpg]] 
The HA must be considered, but in this case it is expected that no more than one server will do off-line in one time. The probability of more than one server going off-line at the same time is quite low. the number of servers could be raised to achieve better HA capabilities, but it wont speed up the process much and there is a reserve in current state.


The cost of this system configuration is: 
Servers type one: 3x 150 USD 
Servers type two: 2x 100 USD 
Load-balancers: 2x 300 USD 
SANs: 2x 220 USD 
DB server: 1x 200 USD 
Total monthly costs: 1890 USD 
The entire solution is very expensive and it might be more cost effective to buy at least some of the devices. Web servers are relatively cheap, so I would start there.

<h2>Case of DoS attack with no mitigation</h2>
In this case the number of used resources is the same as in previous one, but the DoS mitigation devices is not present.

[[File:Lb_result_2_1.jpg]] 
Shows the number of all generated and processed requests. In here it is quite visible that the system was flooded and wasn't able to process all the requests.


[[File:Lb_result_2_2.jpg]] 
The wait time for resource is absolutely unacceptable, basically all the requests would time out before they would even start being processed.


[[File:Lb_result_2_3.jpg]] 
All the web servers are flooded and cant handle the number of incoming requests.


This shows that a DoS attack cant crush the entire system so proper measures must be in place. As is illustrated in previous case.

=Conclusion=

The simulation has identified the optimal number of web servers and other involved components. As expected the costs of such system are quite high. 
Because the simulation was made using SimProcess, there are some limitations as to defining the resource limitations. In here it represents only the delays and number of units available. It would be better to represent the server capabilities by something else than a delay, however it is sufficient. 
The simulation can be modified to find optimal number of servers and other components for a different number of requests than used. In this simulation the optimal numbers are: 3 web servers of type one, 2 web servery of type two, two load-balancers (only due to redundancy, otherwise one would be enough), two SANs (same reason), one DB. The overall monthly cost of such a system is about 1890 USD. 
The simulation has also illustrated the effect of a DoS attack and a possible protection against it. The additional costs for such a protection depends on too many factors. However in some datacenters it is possible to lease a part of the device, that does the mitigation (it could start at 50 USD per month).

=Code=
[[File:Load_balancing.spm]]

=Sources=
http://searchnetworking.techtarget.com/definition/load-balancing 
http://searchstorage.techtarget.com/definition/failover 
http://searchsoftwarequality.techtarget.com/definition/denial-of-service 
http://searchsecurity.techtarget.com/definition/distributed-denial-of-service-attack 
http://cepa.io/devlog/secure-https-load-balancing-with-nginx

Load-balancing

2016-01-17T22:29:33Z

Xtomp36: /* Method */

*'''Project name:''' Load-balancing
*'''Class:''' 4IT496 (WS 2015/2016)
*'''Author:''' Bc. Patrik Tomášek
*'''Model type:''' Discrete-event simulation
*'''Software used:''' SimProcess, trial version

=Problem definition=

A hosting company with its own infrastructure is using so called "load balancing" to distribute the overall load between multiple servers (hardware nodes) and “high-availability” to minimize service down-time.

'''Load-balancing''' 
In the case of web hosting service it means dividing the incoming request between multiple devices (called nodes) based on a set of rules (priority, weight, etc). The simulation is based on nginx (a webserver software) load balancing.
 
[[File:Nginx_load_balancing.png]]
 
''Note: In the picture the load-balacer isn't redundant, therefore HA isn't enabled. The simulation has two load-balances present.''


'''High-avaibility''' 
Ensures that a system or component is operational for desirable time. The solution necessary to provide web hosting consists of many parts, where all of them need to be on-line for the whole to be operational.

To enable HA a provider can use failover and backups. Failover is basically a backup piece of hardware, which ensures that when a component goes off-line another takes it's place. After that it's necessary to load the backup on the component that took over. If there is a SAN (storage area network) implemented than there is no need to load a backup, because failover just uses the same data from one central storage, which is used for all the server nodes. In the simulation a SAN is implemented and failover is taken into consideration.


'''Anti-DoS/DDoS''' 
Denial-of-service (DoS) attack is an incident is witch the targeted service goes down. Distributed denial-of-service means, that more than one system is used to attack a single target. There are more means of possible protection against such attack. Setting up a decent firewall rules might be a good place to start, but it isn't so effective as implementing a device, which can mitigate the attack. A Radware defencePro device is implemented in the simulation.



The goal of the simulation is to find the optimal number of server and other components necessary to enable the mentioned functions (LB, HA, Anti-DoS).


=Method=
Discrete event simulation can be solved using other software than SimProcess, however graphical interface was preferred. It is more user friendly to use GUI to create the simulation than typing it in code. 
There are some limitations present due to use of a trial revision, but this simulation doesn't reached them.

=Model=


The simulation is divided into 4 main processes: 
'''- New requests''' - generation of new requests, further information in Entities section. 
'''- DoS mitigation''' - implementation of anti-dos device. 
'''- Load balancing''' - main process, witch distributes the generated (incoming) request between resources. 
'''- Served request''' - simple dispose of the generated requests. 
[[File:Whole_model.jpg]]



The simulation is set to run in two 9 minutes iterations, using multiple schedules to apply the content changes of a web page (cached/un-cached content). 
Generated request are the highest recorded numbers of the given hosting environment (so called peak). Other possibilities doesn't need to be taken into consideration because the system must be able to handle the peak and with owned hardware, there is not much use for downscaling. Besides the peak might come at odd hour, and starting up an off-line server takes a lot of time. It would be better to use the unallocated server usage for some other calculation, which might be beneficial for the provider. 
'''Time unit used''': seconds 
'''Number of all requests:''' 9000 per second is the maximum recorded in the given environment. The number was divided by 100 for the simulation purposes (so it's 90). 
'''Number of replications:''' 2

== Entities ==
This is a list of all defined and used entities within the simulation. 


'''Requests''' 
The request data are based on real data from an hosting environment, that hosts multiple Magento e-shops (a rather complex system, which uses a lot of hardware). 
The incoming request are categorised because Magento uses caching and indexing. It takes less time to server a cached content than un-cached one. Further more there are multiple schedules used to generate the requests. If new data have been added to the e-shop, that they need to be cached first, witch means more large requests.


'''- Small request''' 
Request of a cached content.

'''- Standard request''' 
Request of a content on storage server, no need to accesses database.

'''- Large request''' 
Request of a content on storage server and database.

'''D-DoS''' 
A large number of requests aiming to cripple the system.


== Resources ==

''This is only a brief summary of resources used, further information follows in processes section of this page.'' 
When it comes to resource price, I take only lease into consideration. Of course it is also possible to purchase the given component. Cloud computing is also not considered.

'''Load-balancer''' 
An important component described earlier on this page. 
These devices are quite expensive. In case of lease, the price can be 300 USD monthly for a sufficient component for this simulation requirements.

'''Web server''' 
There are two types of this resource, each with different capabilities: 
Type 1 - a faster server that goes for 150 USD monthly 
Type 2 - a slower server that goes for 100 USD monthly 

'''Storage server SAN''' 
In the simulation only SAN is used as a name of this resource.
The price of one SAN depends on the configuration (number of HDDs, etc.). A sufficient SAN can be leased for 220 USD monthly.

'''Database server''' 
Database server is quite fast, so in the simulation only one is necessary, however to achieve HA even with DB server a second one should be implemented.
The price of one DB server is about 200 USD monthly.

'''Radware DefensePro''' 
This is an anti-dos component implemented in the simulation. 
The cost of this hardware is extreme, therefore partial lease is used. 
The price depends on the internet connection of the provider, is this case it goes around 50 USD monthly. 


== Processes ==

'''New requests''' 

[[File:Request_generate.jpg]]

In this process new requests are generated according to the following schedule: 
'''Interval:''' 1 second 

{| class="wikitable"
|-
! Entity type
! Most content cached
! Standard content
! New content added
! Sum of normal requests
! Addition DoS requests (red square)
|-
| New small request (green dot)
| Poi(45)
| Poi(25)
| Poi(13.5)
| Poi(90)
| Poi(150)
|-
| New standard request (orange dot)
| Poi(31.5)
| Poi(45)
| Poi(31.5)
| Poi(90)
| Poi(150)
|-
| New large request (black dot)
| Poi(13.5)
| Poi(20)
| Poi(45)
| Poi(90)
| Poi(150)
|-
|}
 


'''DoS mittigation''' 

[[File:Dos_mittigation.jpg]]

This is a simple process, which implements anti-dos device and ensures that the system is not effected by the attack. 
The "Branch" activity simply lets forward only the correct request and the rest is disposed, thus the attack is mitigated. 
Without this device implemented the system would get flooded as later illustrated in results section of this page.


'''Load balancing''' 

[[File:Load balancing.jpg]]


This is the main process of the simulation. Firstly it is expected that about 1 % of incoming request are lost somewhere on the way, so that is what the "Route" activity does. The lost requests are simply disposed. Than the request arrives at Load-balancer, which divides them based on probability. It would be also possible to use least processing or something similar. The load-balancers are designed to handle extreme amount of requests and are very fast, therefore there is only as little delay as 1 ms in the simulation and event that is probably to high. Basically with the number of generated request in the simulation the LB can't be overloaded even in case of DoS attack. However it is a critical piece of hardware, because if it goes down, no request will reach the host, therefore it must be redundant to enable HA (so two LBs are required). 
After this the request reach either Server of type 1 or type 2. Server type one is a delay of Nor(25,3) ms and server type 2 is a delay of Nor(40,3.0) ms, because it is slower. Upon the web server processes the request it is decided where the request goes next based on Entity.type. In case of a small request it goes straight to the next Branch "Response". The standard and large request are send to storage server. The storage server represents another delay an because the system uses SAN it is a common resource for all the web servers. The delay of SAN is defined as Nor(10.0,2.0) ms. Standard request is than send to "Response" activity by "Entity route 2" activity. Large requests are further more send to DB server (the same delay as in the case of storage server) by the same activity. After that they continue to "Response" activity. Branch activity "Response" serves the content to the client, but only 90 % of requests are correct, 7 % are incorrect and client gives up and 3 % are also incorrect but client send a new request.
 


'''Requests served''' 
[[File:Served_requests.jpg]]

A simple dispose of served requests.


=Results=
This section is divided into the following cases.

<h2>Optimal number of servers with DoS attack mitigation present</h2>

The simulation has shown that the optimal number of server necessary to process incoming request properly is 3 of type one and 2 of type two.

[[File:Lb_result_1.jpg]] 
Shows the number of all generated and processed requests.


[[File:Lb_result_2.jpg]] 
The wait time for resource is within the acceptable bounds. If it would be over 100 ms, some improvement should be done.


[[File:Lb_result_3.jpg]] 
The HA must be considered, but in this case it is expected that no more than one server will do off-line in one time. The probability of more than one server going off-line at the same time is quite low. the number of servers could be raised to achieve better HA capabilities, but it wont speed up the process much and there is a reserve in current state.


The cost of this system configuration is: 
Servers type one: 3x 150 USD 
Servers type two: 2x 100 USD 
Load-balancers: 2x 300 USD 
SANs: 2x 220 USD 
DB server: 1x 200 USD 
Total monthly costs: 1890 USD 
The entire solution is very expensive and it might be more cost effective to buy at least some of the devices. Web servers are relatively cheap, so I would start there.

<h2>Case of DoS attack with no mitigation</h2>
In this case the number of used resources is the same as in previous one, but the DoS mitigation devices is not present.

[[File:Lb_result_2_1.jpg]] 
Shows the number of all generated and processed requests. In here it is quite visible that the system was flooded and wasn't able to process all the requests.


[[File:Lb_result_2_2.jpg]] 
The wait time for resource is absolutely unacceptable, basically all the requests would time out before they would even start being processed.


[[File:Lb_result_2_3.jpg]] 
All the web servers are flooded and cant handle the number of incoming requests.


This shows that a DoS attack cant crush the entire system so proper measures must be in place. As is illustrated in previous case.

=Conclusion=

The simulation has identified the optimal number of web servers and other involved components. As expected the costs of such system are quite high. 
Because the simulation was made using SimProcess, there are some limitations as to defining the resource limitations. In here it represents only the delays and number of units available. It would be better to represent the server capabilities by something else than a delay, however it is sufficient. 
The simulation can be modified to find optimal number of servers and other components for a different number of requests than used. In this simulation the optimal numbers are: 3 web servers of type one, 2 web servery of type two, two load-balancers (only due to redundancy, otherwise one would be enough), two SANs (same reason), one DB. The overall monthly cost of such a system is about 1890 USD. 
The simulation has also illustrated the effect of a DoS attack and a possible protection against it. The additional costs for such a protection depends on too many factors. However in some datacenters it is possible to lease a part of the device, that does the mitigation (it could start at 50 USD per month).

=Code=
[[File:Load_balancing.spm]]

=Sources=
http://searchnetworking.techtarget.com/definition/load-balancing 
http://searchstorage.techtarget.com/definition/failover 
http://searchsoftwarequality.techtarget.com/definition/denial-of-service 
http://searchsecurity.techtarget.com/definition/distributed-denial-of-service-attack 
http://cepa.io/devlog/secure-https-load-balancing-with-nginx