March 2017 MARKET SEGMENTATION A GUIDE TO CONDUCTING SEGMENTATION RESEARCH
Market segmentation is a popular technique people likely to try this new technology. A in marketing, marketing research and limited segmentation scheme such as this is strategic planning. Market segmentation justified because the first people to try a new identifies unique groups of consumers in a product are rare, relatively homogeneous as market for various end-goals, including: early adopters/innovators, and do not require di erentiated marketing campaigns. 1 Better understanding of consumers Further segmentation would be costly and Tailoring of marketing would require committing resources that 2 communication messages could be more e ectively used elsewhere. New product development/line In this stage of the category life cycle further 3 extensions/re-branding segmentation is not justified. Identifying potential economic/ As a product category matures, however, 4 volumetric opportunities of new features are added to existing products, the segments the product gains distribution, more people The purpose of this document is to outline become aware of the product, and buyers when it is appropriate to consider segmentation become more savvy about the product and and the basic steps for executing its benefits. Likewise, competition increases. segmentation research. At this point, the seller must focus on more than selling a “new” product. The seller must WHEN TO CONSIDER SEGMENTATION also identify and meet the di ering needs of the market. The need for market segmentation can be shown by examining the category life cycle. As needs or desires expand into various stages In the early stages of a new product category, of heterogeneity, the seller is faced with there are few brands and few buyers. Market continuing to consider all buyers as similar or segmentation at this stage is simply separating adopt some type of market segmentation. those likely to try the new product from those The purpose of segmentation research, then, less likely. Media campaigns, advertising copy, is to find those unique subsets of the market and distribution channels are developed with that can be identified and treated as a group. this “likely to buy” group in mind. For example, The goal is to identify groups such that when the VCR was first introduced, the buying behavior within a group is similar marketing e orts were targeted at those (segment homogeneity) and the behavior
1
between groups is di erent. If the buyers have homogeneous needs, a company is best served by not segmenting the market. This is the case with most new (or innovative) products and some commodity products. Virtually every market, however, can benefit from a segmentation approach.
BASIC STEPS FOR which respondents are to be grouped based SEGMENTATION RESEARCH on age and household income. An a priori segmentation scheme could use the In deciding how to identify segments, the following categories shown above. researcher must answer four questions: In this scheme, each respondent could be What is the key objective of placed in the proper category as soon as the segmentation? these two pieces of information were known. What method should be used to segment the market? Post hoc segmentation, on the other hand, requires only that the variables on which the What variables, or basis, should be segmentation will be based (but not the used to segment the market? segment definitions) be identified. The How many segments should segments are then determined by using an be formed? analytical technique (such as cluster analysis) after all the data is collected. For example, if 1. Objective post hoc segments were to be formed on Frequently, the initial decision to conduct psychographics, the psychographic scores segmentation research is based on the for all respondents would be cluster analyzed. desire to better understand consumers. The segments would not be known until Though this is a key starting point, it is after running cluster analysis. This allows important to subsequently detail the groups to be formed based on "natural" strategic objectives to be met by this breaks in the data. This method is sometimes better understanding. This is the time to referred to as response-based. consider how the segments will be used The post hoc benefit segmentation approach once they have been identified. Taking the has become the preferred choice in most time to do this at the start of the process large “baseline” segmentation studies. By is essential for successful knowing what benefits buyers in a specific segmentation research. segment are looking to receive from a product, 2. Method the marketing mix is more clearly guided to appeal exactly to that target market. The method of segmentation refers to when the segments are defined. There are two 3. Basis methods of segmentation. After determining the Method of segmentation, They are a priori and post hoc. the researcher must determine what characteristics to use to identify the segments. Segmentation requires that respondents be These characteristics are referred to as the grouped based on some set of variables that basis of segmentation. Unlike the method are identified before data collection. In a question which had two possible answers, priori segmentation, not only are the the basis question is limited only by the variables identified, but so are the segment definitions. For example, consider a study in
researcher’s imagination. The following table illustrates some of the most frequently used bases of segmentation. Each of the major bases are discussed below:
Geographic Geographic segmentation calls for dividing the market into geographic regions. This allows the company to operate in only a few regions, because of limited demand or resources, or the company can tailor some element of the marketing mix by region. Food products are frequently varied by region. Either some products are only o ered in certain markets, or the di ering products are sold in the same container with the same advertising and promotion in di erent geographies.
Demographic Historically, the most common bases of segmentation have been demographic in nature – women between 18 and 49, for example. These schemes have frequently been motivated from a shopping or media behavior perspective and not one of brand or category choice behavior. While these schemes might adequately identify the frequent purchasers of a product category, they generally don’t fit the definition of segmentation as outlined above – dividing consumers by product choice behavior. Income, education, ethnicity, and occupation are all frequently used demographic segmentation variables.
Attitudinal Attitudinal variables are used based on the idea that opinions, traits, interests, and lifestyles influence buying behavior. Put another way, people with similar psychographic profiles should respond similarly to various marketing mixes. It is not clear that psychographics directly influence brand choice but are likely related to usage in several categories. Their application to market segmentation is well established.
Product Usage Product usage variables are also used to
segment markets based on category usage, are derived based on causal factors relating to product usage, or brand loyalty. It is a the purchase process. People buy products, and common practice to segment the market decide between di erent brands, because of into users and non-users and also by level of the perceived unique benefits that one use (heavy vs. light). product can provide over another, and the relative importance of those benefits. Benefit Decision Process segmentation partitions the market into People decide which product to buy in segments based on benefits sought. The di erent ways. The influence that people result is that the people in one segment are exercise over decisions, as well as their brand seeking to realize some rather homogeneous loyalty, variety seeking behavior, information set of benefits with a particular product. search, distribution channel used and Those familiar with conjoint analysis can decision making units (centralized versus readily see the applicability for conjoint decentralized) all impact the brand that they to measure the relative importances will purchase. While this focus is more of the benefits sought. prevalent in business-to-business settings, certain elements are perfectly applicable The post hoc benefit segmentation approach to consumer segmentation as well. has become the preferred choice in most large “baseline” segmentation studies. By knowing Firmographic what benefits buyers in a specific segment The firmographic variables of SIC, revenue are looking to receive from a product, the and number of employees are standard bases marketing mix is more clearly guided to for industrial segmentation. Firmographic appeal exactly to that target market. segmentation is based on the idea that 4. Number of Segments companies can be targeted based simply on what industry they are in or how large Another question that must be answered in they are. This approach has been criticized any segmentation study is how many segments as largely ine ective. to use. The decision on how many segments has considerations from both statistical and Adoption Propensity managerial viewpoints. Segments can almost As shown when introducing the idea of always be made more homogeneous by segmentation, new product introductions increasing the number of segments. This could can likely be made more e cient by targeting continue until the number of segments becomes the early adopters of a product category. unmanageable. One practical guideline is Segmentation based on adoption propensity that unless the market can be broken into might be based on the propensity for the at least three groups, the advantages of specific product or category, or it might be segmenting are not cost e ective over a based on a buyer’s characteristic or propensity general market approach; and attempting to to adopt new product across several categories. market to more than seven segments is not This approach might even involve opinion practical for most companies. leadership – people whose opinions are sought There are obviously many possible ways to even though they might not be early adopters. segment a market. There is no single right Needs-Based or Benefit Segmentation answer and each situation can produce a di erent recommended “best” solution. What Needs-based or benefit segmentation is should be obvious is that although substantially di erent from the methods segmentation research involves substantial mentioned above. Instead of relying on scientific processes, it also relies inescapably descriptive factors (like region, age, income, on managerial judgment and the brand loyalty, or product usage), segments art of segmentation.
HOW TO USE MARKET SEGMENTS? competitive positioning, and market/industry dynamics. The basic approach, especially in To be most useful, the market segments just benefit segmentation, is to evaluate each identified should be tracked in all future segment’s desired product and evaluate research. To do this, researchers must develop whether current or new products can meet a way to make the segments identifiable and that segment’s needs without deteriorating meaningful so that they can be incorporated current franchise positioning. For example, in future e orts. This happens through many niche products which can command a segment profiles or descriptions. Segment price premium in their niche would require descriptions are useful for two primary reasons. price parity or even discounts when outside Most large segmentation studies of their niche. In this case, appealing to use variables that are not easily segments beyond the current niche is likely observable as the basis of unprofitable. This consideration of targeting segmentation. Measurements like must go beyond just the market’s desired large psychographic batteries, product specifications, but must also include part-worth utilities from conjoint and their willingness to consider a particular product preferences are frequently brand and distribution channels. used basis variables. Instead, the WHAT DOES SEGMENTATION segment profiles can be used to REQUIRE? identify the segments. E ective segmentation requires that truly Profiling the segments can provide di erent groups be identified. They must be: a validity check for the segmentation results themselves. The profiling Large enough to be economically stage can assist in the identification sustainable (size) of meaningless or artificial segments. If the profiles don't make sense Di erent enough to justify the from a managerial perspective, increased cost of segmentation the segmentation scheme must (di erent) be reconsidered. Similar enough within each segment Typical profiling variables are demographics, to be treated as one group psychographics, or other descriptive (homogeneous) variables. There are a number of proven Identifiable through segment research techniques that assist in profiling profiling (actionable) the segments. The methods range from simple cross-tabs to complex multivariate Accessible through unique analysis procedures. Examples include media/distribution (accessible) multiple discriminant analysis, logistic regression, correspondence analysis, Stable enough to exist over some and CHAID. planning horizon (stable)
WHICH SEGMENTS DO WE TARGET? Each of these six requirements must be met to justify a segmentation approach. These Once segments are identified and profiled, requirements are each discussed in turn. there is one more inescapable question: Do we target all of these segments or are our Size products suited for just a subset of the The segments must be large enough to segments? The answer to this question has sustain the additional cost incurred by many considerations: available resources, segmentation (like developing specialized current company image and its portability, communication messages). Segment size is
a relative question based on the size of the instance, there would be little trouble company, the size of the market, and the assigning individuals to the proper segment. buying power of each customer. Boeing, If however, a more sophisticated scheme is for example, would be able to define each used to identify segments, the researcher potential customer as a separate segment must be able to assign observations to and tailor their marketing mix to each segments independent of the basis variables. buyer's individual needs. A manufacturer Segments from post hoc assignment are of breakfast cereal, however, must segment formed on basis variables that are di cult the market on a more general basis. to measure--like attitudes or attribute importances. If the segments from a post Di erent hoc segmentation scheme cannot be The individual segments must be di erent profiled, the segments will be di cult to from each other. If all buyers in the market carry forward into future studies. behave basically the same, there is no need Accessible to segment the market. Likewise, if segments are found in which the behaviors across The segments must also be reachable. To groups are basically indistinguishable, the the extent that a segment exists, but cannot segmentation scheme is not accomplishing be educated about a product's potential what it should be. Markets for non- benefits by advertising, point of purchase, di erentiated, or commodity products or another method, it is of very little use to would not benefit from a segmented the company and of no more value than approach. A manufacturer of commodity the unsegmented market. products can more e ciently reach the entire market with one marketing strategy Stable than with several. There are, however, few Stability over time is also a reasonable markets that are absolutely not segmentable. concern for segmentation studies. The That is, almost every market has identifiable question is not so much that one person groups that behave di erently and by stays in the same segment all the time, but focusing on these di erences, companies that over time, the same basic segments can increase their marketing e ectiveness. exist, in about the same size, with the same Homogeneous basic profiles. Just as the segments need to be di erent, WHAT MAKES SEGMENTATION those potential buyers in the same segment RESEARCH SUCCESSFUL? need to be similar. That is, they must behave As pointed out above, almost any marketing similarly. This is necessary to treat the e ort can be made more e ective with segment as one entity. market segmentation, and almost any Actionable marketing research can be made more informative by analyzing the results at the Segments must be profiled so that potential segment level. buyers can be placed in a segment, based on easily measurable variables. This allows the While most people can clearly see the researcher to use the segmentation scheme benefits of segmentation, many are hesitant in future research without having to recreate to commit the time and e ort necessary to the segmentation scheme. These profiles perform a good segmentation study. Many could, for a priori segmentation schemes, people are tempted to perform a be the basis variables themselves. If rudimentary segmentation study hoping that segments were based on age and number the outcome will be as if the research design of children, or geographic region, for had been fully, and carefully, developed. This
fear has prompted some to refer to the “generic segmentation study, one done mindlessly and mechanically, following specific steps by rote under the assumption that clear-cut and insightful segments will somehow magically pop out of the welter of statistical output generated."1 The keys to successful segmentation research are:
Clearly defined goals and objectives
A substantial amount of up-front work
A thorough examination of segmentation dimensions
Well developed plans for follow- through activities (use what you have learned) Market segmentation is an extremely powerful strategic marketing tool. It must be treated as a strategic tool, though, and involve all levels of management and all future marketing decisions. When well considered, well planned and well executed, segmentation research can provide insights into your market that would otherwise go unnoticed and can provide increased e ciencies in all marketing e orts: product design, communication, distribution, and pricing. Market segmentation is a popular technique people likely to try this new technology. A in marketing, marketing research and limited segmentation scheme such as this is strategic planning. Market segmentation justified because the first people to try a new identifies unique groups of consumers in a product are rare, relatively homogeneous as market for various end-goals, including: early adopters/innovators, and do not require di erentiated marketing campaigns. Better understanding of consumers Further segmentation would be costly and Tailoring of marketing would require committing resources that communication messages could be more e ectively used elsewhere. New product development/line In this stage of the category life cycle further extensions/re-branding segmentation is not justified. Identifying potential economic/ As a product category matures, however, volumetric opportunities of new features are added to existing products, the segments the product gains distribution, more people The purpose of this document is to outline become aware of the product, and buyers when it is appropriate to consider segmentation become more savvy about the product and and the basic steps for executing its benefits. Likewise, competition increases. segmentation research. At this point, the seller must focus on more than selling a “new” product. The seller must WHEN TO CONSIDER SEGMENTATION also identify and meet the di ering needs of the market. The need for market segmentation can be shown by examining the category life cycle. As needs or desires expand into various stages In the early stages of a new product category, of heterogeneity, the seller is faced with there are few brands and few buyers. Market continuing to consider all buyers as similar or segmentation at this stage is simply separating adopt some type of market segmentation. those likely to try the new product from those The purpose of segmentation research, then, less likely. Media campaigns, advertising copy, is to find those unique subsets of the market and distribution channels are developed with that can be identified and treated as a group. this “likely to buy” group in mind. For example, The goal is to identify groups such that when the VCR was first introduced, the buying behavior within a group is similar marketing e orts were targeted at those (segment homogeneity) and the behavior
MARKET SEGMENTATION | MARKETVISION RESEARCH
between groups is di erent. HH Income If the buyers have homogeneous needs, a < $25k <$25k - $75k > $75k company is best served by not segmenting A 18-34 the market. This is the case with most new (or innovative) products and some commodity G 35-54 products. Virtually every market, however, E 55+ can benefit from a segmentation approach.
BASIC STEPS FOR which respondents are to be grouped based SEGMENTATION RESEARCH on age and household income. An a priori segmentation scheme could use the In deciding how to identify segments, the following categories shown above. researcher must answer four questions: In this scheme, each respondent could be What is the key objective of placed in the proper category as soon as 1 the segmentation? these two pieces of information were known. What method should be used 2 to segment the market? Post hoc segmentation, on the other hand, requires only that the variables on which the What variables, or basis, should be segmentation will be based (but not the 3 used to segment the market? segment definitions) be identified. The How many segments should 4 segments are then determined by using an be formed? analytical technique (such as cluster analysis) after all the data is collected. For example, if 1. Objective post hoc segments were to be formed on Frequently, the initial decision to conduct psychographics, the psychographic scores segmentation research is based on the for all respondents would be cluster analyzed. desire to better understand consumers. The segments would not be known until Though this is a key starting point, it is after running cluster analysis. This allows important to subsequently detail the groups to be formed based on "natural" strategic objectives to be met by this breaks in the data. This method is sometimes better understanding. This is the time to referred to as response-based. consider how the segments will be used The post hoc benefit segmentation approach once they have been identified. Taking the has become the preferred choice in most time to do this at the start of the process large “baseline” segmentation studies. By is essential for successful knowing what benefits buyers in a specific segmentation research. segment are looking to receive from a product, 2. Method the marketing mix is more clearly guided to appeal exactly to that target market. The method of segmentation refers to when the segments are defined. There are two 3. Basis methods of segmentation. After determining the Method of segmentation, They are a priori and post hoc. the researcher must determine what characteristics to use to identify the segments. Segmentation requires that respondents be These characteristics are referred to as the grouped based on some set of variables that basis of segmentation. Unlike the method are identified before data collection. In a question which had two possible answers, priori segmentation, not only are the the basis question is limited only by the variables identified, but so are the segment definitions. For example, consider a study in
2
researcher’s imagination. The following table illustrates some of the most frequently used bases of segmentation. Each of the major bases are discussed below:
Geographic Geographic segmentation calls for dividing the market into geographic regions. This allows the company to operate in only a few regions, because of limited demand or resources, or the company can tailor some element of the marketing mix by region. Food products are frequently varied by region. Either some products are only o ered in certain markets, or the di ering products are sold in the same container with the same advertising and promotion in di erent geographies.
Demographic Historically, the most common bases of segmentation have been demographic in nature – women between 18 and 49, for example. These schemes have frequently been motivated from a shopping or media behavior perspective and not one of brand or category choice behavior. While these schemes might adequately identify the frequent purchasers of a product category, they generally don’t fit the definition of segmentation as outlined above – dividing consumers by product choice behavior. Income, education, ethnicity, and occupation are all frequently used demographic segmentation variables.
Attitudinal Attitudinal variables are used based on the idea that opinions, traits, interests, and lifestyles influence buying behavior. Put another way, people with similar psychographic profiles should respond similarly to various marketing mixes. It is not clear that psychographics directly influence brand choice but are likely related to usage in several categories. Their application to market segmentation is well established.
Product Usage Product usage variables are also used to
segment markets based on category usage, are derived based on causal factors relating to product usage, or brand loyalty. It is a the purchase process. People buy products, and common practice to segment the market decide between di erent brands, because of into users and non-users and also by level of the perceived unique benefits that one use (heavy vs. light). product can provide over another, and the relative importance of those benefits. Benefit Decision Process segmentation partitions the market into People decide which product to buy in segments based on benefits sought. The di erent ways. The influence that people result is that the people in one segment are exercise over decisions, as well as their brand seeking to realize some rather homogeneous loyalty, variety seeking behavior, information set of benefits with a particular product. search, distribution channel used and Those familiar with conjoint analysis can decision making units (centralized versus readily see the applicability for conjoint decentralized) all impact the brand that they to measure the relative importances will purchase. While this focus is more of the benefits sought. prevalent in business-to-business settings, certain elements are perfectly applicable The post hoc benefit segmentation approach to consumer segmentation as well. has become the preferred choice in most large “baseline” segmentation studies. By knowing Firmographic what benefits buyers in a specific segment The firmographic variables of SIC, revenue are looking to receive from a product, the and number of employees are standard bases marketing mix is more clearly guided to for industrial segmentation. Firmographic appeal exactly to that target market. segmentation is based on the idea that 4. Number of Segments companies can be targeted based simply on what industry they are in or how large Another question that must be answered in they are. This approach has been criticized any segmentation study is how many segments as largely ine ective. to use. The decision on how many segments has considerations from both statistical and Adoption Propensity managerial viewpoints. Segments can almost As shown when introducing the idea of always be made more homogeneous by segmentation, new product introductions increasing the number of segments. This could can likely be made more e cient by targeting continue until the number of segments becomes the early adopters of a product category. unmanageable. One practical guideline is Segmentation based on adoption propensity that unless the market can be broken into might be based on the propensity for the at least three groups, the advantages of specific product or category, or it might be segmenting are not cost e ective over a based on a buyer’s characteristic or propensity general market approach; and attempting to to adopt new product across several categories. market to more than seven segments is not This approach might even involve opinion practical for most companies. leadership – people whose opinions are sought There are obviously many possible ways to even though they might not be early adopters. segment a market. There is no single right Needs-Based or Benefit Segmentation answer and each situation can produce a di erent recommended “best” solution. What Needs-based or benefit segmentation is should be obvious is that although substantially di erent from the methods segmentation research involves substantial mentioned above. Instead of relying on scientific processes, it also relies inescapably descriptive factors (like region, age, income, on managerial judgment and the brand loyalty, or product usage), segments art of segmentation.
HOW TO USE MARKET SEGMENTS? competitive positioning, and market/industry dynamics. The basic approach, especially in To be most useful, the market segments just benefit segmentation, is to evaluate each identified should be tracked in all future segment’s desired product and evaluate research. To do this, researchers must develop whether current or new products can meet a way to make the segments identifiable and that segment’s needs without deteriorating meaningful so that they can be incorporated current franchise positioning. For example, in future e orts. This happens through many niche products which can command a segment profiles or descriptions. Segment price premium in their niche would require descriptions are useful for two primary reasons. price parity or even discounts when outside Most large segmentation studies of their niche. In this case, appealing to use variables that are not easily segments beyond the current niche is likely observable as the basis of unprofitable. This consideration of targeting segmentation. Measurements like must go beyond just the market’s desired large psychographic batteries, product specifications, but must also include part-worth utilities from conjoint and their willingness to consider a particular product preferences are frequently brand and distribution channels. used basis variables. Instead, the WHAT DOES SEGMENTATION segment profiles can be used to REQUIRE? identify the segments. E ective segmentation requires that truly Profiling the segments can provide di erent groups be identified. They must be: a validity check for the segmentation results themselves. The profiling Large enough to be economically stage can assist in the identification sustainable (size) of meaningless or artificial segments. If the profiles don't make sense Di erent enough to justify the from a managerial perspective, increased cost of segmentation the segmentation scheme must (di erent) be reconsidered. Similar enough within each segment Typical profiling variables are demographics, to be treated as one group psychographics, or other descriptive (homogeneous) variables. There are a number of proven Identifiable through segment research techniques that assist in profiling profiling (actionable) the segments. The methods range from simple cross-tabs to complex multivariate Accessible through unique analysis procedures. Examples include media/distribution (accessible) multiple discriminant analysis, logistic regression, correspondence analysis, Stable enough to exist over some and CHAID. planning horizon (stable)
WHICH SEGMENTS DO WE TARGET? Each of these six requirements must be met to justify a segmentation approach. These Once segments are identified and profiled, requirements are each discussed in turn. there is one more inescapable question: Do we target all of these segments or are our Size products suited for just a subset of the The segments must be large enough to segments? The answer to this question has sustain the additional cost incurred by many considerations: available resources, segmentation (like developing specialized current company image and its portability, communication messages). Segment size is
a relative question based on the size of the instance, there would be little trouble company, the size of the market, and the assigning individuals to the proper segment. buying power of each customer. Boeing, If however, a more sophisticated scheme is for example, would be able to define each used to identify segments, the researcher potential customer as a separate segment must be able to assign observations to and tailor their marketing mix to each segments independent of the basis variables. buyer's individual needs. A manufacturer Segments from post hoc assignment are of breakfast cereal, however, must segment formed on basis variables that are di cult the market on a more general basis. to measure--like attitudes or attribute importances. If the segments from a post Di erent hoc segmentation scheme cannot be The individual segments must be di erent profiled, the segments will be di cult to from each other. If all buyers in the market carry forward into future studies. behave basically the same, there is no need Accessible to segment the market. Likewise, if segments are found in which the behaviors across The segments must also be reachable. To groups are basically indistinguishable, the the extent that a segment exists, but cannot segmentation scheme is not accomplishing be educated about a product's potential what it should be. Markets for non- benefits by advertising, point of purchase, di erentiated, or commodity products or another method, it is of very little use to would not benefit from a segmented the company and of no more value than approach. A manufacturer of commodity the unsegmented market. products can more e ciently reach the entire market with one marketing strategy Stable than with several. There are, however, few Stability over time is also a reasonable markets that are absolutely not segmentable. concern for segmentation studies. The That is, almost every market has identifiable question is not so much that one person groups that behave di erently and by stays in the same segment all the time, but focusing on these di erences, companies that over time, the same basic segments can increase their marketing e ectiveness. exist, in about the same size, with the same Homogeneous basic profiles. Just as the segments need to be di erent, WHAT MAKES SEGMENTATION those potential buyers in the same segment RESEARCH SUCCESSFUL? need to be similar. That is, they must behave As pointed out above, almost any marketing similarly. This is necessary to treat the e ort can be made more e ective with segment as one entity. market segmentation, and almost any Actionable marketing research can be made more informative by analyzing the results at the Segments must be profiled so that potential segment level. buyers can be placed in a segment, based on easily measurable variables. This allows the While most people can clearly see the researcher to use the segmentation scheme benefits of segmentation, many are hesitant in future research without having to recreate to commit the time and e ort necessary to the segmentation scheme. These profiles perform a good segmentation study. Many could, for a priori segmentation schemes, people are tempted to perform a be the basis variables themselves. If rudimentary segmentation study hoping that segments were based on age and number the outcome will be as if the research design of children, or geographic region, for had been fully, and carefully, developed. This
fear has prompted some to refer to the “generic segmentation study, one done mindlessly and mechanically, following specific steps by rote under the assumption that clear-cut and insightful segments will somehow magically pop out of the welter of statistical output generated."1 The keys to successful segmentation research are:
Clearly defined goals and objectives
A substantial amount of up-front work
A thorough examination of segmentation dimensions
Well developed plans for follow- through activities (use what you have learned) Market segmentation is an extremely powerful strategic marketing tool. It must be treated as a strategic tool, though, and involve all levels of management and all future marketing decisions. When well considered, well planned and well executed, segmentation research can provide insights into your market that would otherwise go unnoticed and can provide increased e ciencies in all marketing e orts: product design, communication, distribution, and pricing. Market segmentation is a popular technique people likely to try this new technology. A in marketing, marketing research and limited segmentation scheme such as this is strategic planning. Market segmentation justified because the first people to try a new identifies unique groups of consumers in a product are rare, relatively homogeneous as market for various end-goals, including: early adopters/innovators, and do not require di erentiated marketing campaigns. Better understanding of consumers Further segmentation would be costly and Tailoring of marketing would require committing resources that communication messages could be more e ectively used elsewhere. New product development/line In this stage of the category life cycle further extensions/re-branding segmentation is not justified. Identifying potential economic/ As a product category matures, however, volumetric opportunities of new features are added to existing products, the segments the product gains distribution, more people The purpose of this document is to outline become aware of the product, and buyers when it is appropriate to consider segmentation become more savvy about the product and and the basic steps for executing its benefits. Likewise, competition increases. segmentation research. At this point, the seller must focus on more than selling a “new” product. The seller must WHEN TO CONSIDER SEGMENTATION also identify and meet the di ering needs of the market. The need for market segmentation can be shown by examining the category life cycle. As needs or desires expand into various stages In the early stages of a new product category, of heterogeneity, the seller is faced with there are few brands and few buyers. Market continuing to consider all buyers as similar or segmentation at this stage is simply separating adopt some type of market segmentation. those likely to try the new product from those The purpose of segmentation research, then, less likely. Media campaigns, advertising copy, is to find those unique subsets of the market and distribution channels are developed with that can be identified and treated as a group. this “likely to buy” group in mind. For example, The goal is to identify groups such that when the VCR was first introduced, the buying behavior within a group is similar marketing e orts were targeted at those (segment homogeneity) and the behavior
between groups is di erent. If the buyers have homogeneous needs, a company is best served by not segmenting the market. This is the case with most new (or innovative) products and some commodity products. Virtually every market, however, can benefit from a segmentation approach.
BASIC STEPS FOR which respondents are to be grouped based SEGMENTATION RESEARCH on age and household income. An a priori segmentation scheme could use the In deciding how to identify segments, the following categories shown above. researcher must answer four questions: In this scheme, each respondent could be What is the key objective of placed in the proper category as soon as the segmentation? these two pieces of information were known. What method should be used to segment the market? Post hoc segmentation, on the other hand, requires only that the variables on which the What variables, or basis, should be segmentation will be based (but not the used to segment the market? segment definitions) be identified. The How many segments should segments are then determined by using an be formed? analytical technique (such as cluster analysis) after all the data is collected. For example, if 1. Objective post hoc segments were to be formed on Frequently, the initial decision to conduct psychographics, the psychographic scores segmentation research is based on the for all respondents would be cluster analyzed. desire to better understand consumers. The segments would not be known until Though this is a key starting point, it is after running cluster analysis. This allows important to subsequently detail the groups to be formed based on "natural" strategic objectives to be met by this breaks in the data. This method is sometimes better understanding. This is the time to referred to as response-based. consider how the segments will be used The post hoc benefit segmentation approach once they have been identified. Taking the has become the preferred choice in most time to do this at the start of the process large “baseline” segmentation studies. By is essential for successful knowing what benefits buyers in a specific segmentation research. segment are looking to receive from a product, 2. Method the marketing mix is more clearly guided to appeal exactly to that target market. The method of segmentation refers to when the segments are defined. There are two 3. Basis methods of segmentation. After determining the Method of segmentation, They are a priori and post hoc. the researcher must determine what characteristics to use to identify the segments. Segmentation requires that respondents be These characteristics are referred to as the grouped based on some set of variables that basis of segmentation. Unlike the method are identified before data collection. In a question which had two possible answers, priori segmentation, not only are the the basis question is limited only by the variables identified, but so are the segment definitions. For example, consider a study in
MARKET SEGMENTATION | MARKETVISION RESEARCH researcher’s imagination. The following table illustrates some of the most frequently used bases of segmentation. Each of the major bases are discussed below: Common Basis Variables GEOGRAPHIC Geographic City/Country/ Geographic segmentation calls for dividing Region Urban/Rural Size the market into geographic regions. This allows the company to operate in only a DEMOGRAPHIC few regions, because of limited demand or Age Gender Income Education resources, or the company can tailor some Presence Occupation Ethnicity element of the marketing mix by region. Food of Children products are frequently varied by region. Either ATTITUDINAL/PSYCHOGRAPHICS some products are only o ered in certain markets, or the di ering products are sold in Lifestyles Interest Opinions Values the same container with the same advertising and promotion in di erent geographies. PRODUCT USAGE OR KNOWLEDGE Demographic Frequency Brand loyalty Attitude toward of use product Historically, the most common bases of Level of Usage segmentation have been demographic in product patterns End use knowledge nature – women between 18 and 49, for example. These schemes have frequently Involvement Purchase with category volume been motivated from a shopping or media behavior perspective and not one of brand DECISION PROCESS or category choice behavior. While these schemes might adequately identify the Decisions Shopping making unit pattern Media use frequent purchasers of a product category, they generally don’t fit the definition of Information Distribution segmentation as outlined above – dividing search channel Buyer profile consumers by product choice behavior. Income, education, ethnicity, and occupation ADOPTION PROPENSITY are all frequently used demographic Interest in Opinion Likelihood of segmentation variables. new products leadership buying new products Attitudinal Attitudinal variables are used based on FIRMOGRAPHIC the idea that opinions, traits, interests, Buying Size and lifestyles influence buying behavior. SIC/Industry (employees/ Put another way, people with similar centralization revenue/worksites) psychographic profiles should respond similarly to various marketing mixes. It is not NEEDS/PRODUCT BENEFITS clear that psychographics directly influence brand choice but are likely related to usage Needs to Perception in several categories. Their application to be filled Expectations of product market segmentation is well established. Product Satisfaction preference Product Usage Product usage variables are also used to
3
segment markets based on category usage, are derived based on causal factors relating to product usage, or brand loyalty. It is a the purchase process. People buy products, and common practice to segment the market decide between di erent brands, because of into users and non-users and also by level of the perceived unique benefits that one use (heavy vs. light). product can provide over another, and the relative importance of those benefits. Benefit Decision Process segmentation partitions the market into People decide which product to buy in segments based on benefits sought. The di erent ways. The influence that people result is that the people in one segment are exercise over decisions, as well as their brand seeking to realize some rather homogeneous loyalty, variety seeking behavior, information set of benefits with a particular product. search, distribution channel used and Those familiar with conjoint analysis can decision making units (centralized versus readily see the applicability for conjoint decentralized) all impact the brand that they to measure the relative importances will purchase. While this focus is more of the benefits sought. prevalent in business-to-business settings, certain elements are perfectly applicable The post hoc benefit segmentation approach to consumer segmentation as well. has become the preferred choice in most large “baseline” segmentation studies. By knowing Firmographic what benefits buyers in a specific segment The firmographic variables of SIC, revenue are looking to receive from a product, the and number of employees are standard bases marketing mix is more clearly guided to for industrial segmentation. Firmographic appeal exactly to that target market. segmentation is based on the idea that 4. Number of Segments companies can be targeted based simply on what industry they are in or how large Another question that must be answered in they are. This approach has been criticized any segmentation study is how many segments as largely ine ective. to use. The decision on how many segments has considerations from both statistical and Adoption Propensity managerial viewpoints. Segments can almost As shown when introducing the idea of always be made more homogeneous by segmentation, new product introductions increasing the number of segments. This could can likely be made more e cient by targeting continue until the number of segments becomes the early adopters of a product category. unmanageable. One practical guideline is Segmentation based on adoption propensity that unless the market can be broken into might be based on the propensity for the at least three groups, the advantages of specific product or category, or it might be segmenting are not cost e ective over a based on a buyer’s characteristic or propensity general market approach; and attempting to to adopt new product across several categories. market to more than seven segments is not This approach might even involve opinion practical for most companies. leadership – people whose opinions are sought There are obviously many possible ways to even though they might not be early adopters. segment a market. There is no single right Needs-Based or Benefit Segmentation answer and each situation can produce a di erent recommended “best” solution. What Needs-based or benefit segmentation is should be obvious is that although substantially di erent from the methods segmentation research involves substantial mentioned above. Instead of relying on scientific processes, it also relies inescapably descriptive factors (like region, age, income, on managerial judgment and the brand loyalty, or product usage), segments art of segmentation.
HOW TO USE MARKET SEGMENTS? competitive positioning, and market/industry dynamics. The basic approach, especially in To be most useful, the market segments just benefit segmentation, is to evaluate each identified should be tracked in all future segment’s desired product and evaluate research. To do this, researchers must develop whether current or new products can meet a way to make the segments identifiable and that segment’s needs without deteriorating meaningful so that they can be incorporated current franchise positioning. For example, in future e orts. This happens through many niche products which can command a segment profiles or descriptions. Segment price premium in their niche would require descriptions are useful for two primary reasons. price parity or even discounts when outside Most large segmentation studies of their niche. In this case, appealing to use variables that are not easily segments beyond the current niche is likely observable as the basis of unprofitable. This consideration of targeting segmentation. Measurements like must go beyond just the market’s desired large psychographic batteries, product specifications, but must also include part-worth utilities from conjoint and their willingness to consider a particular product preferences are frequently brand and distribution channels. used basis variables. Instead, the WHAT DOES SEGMENTATION segment profiles can be used to REQUIRE? identify the segments. E ective segmentation requires that truly Profiling the segments can provide di erent groups be identified. They must be: a validity check for the segmentation results themselves. The profiling Large enough to be economically stage can assist in the identification sustainable (size) of meaningless or artificial segments. If the profiles don't make sense Di erent enough to justify the from a managerial perspective, increased cost of segmentation the segmentation scheme must (di erent) be reconsidered. Similar enough within each segment Typical profiling variables are demographics, to be treated as one group psychographics, or other descriptive (homogeneous) variables. There are a number of proven Identifiable through segment research techniques that assist in profiling profiling (actionable) the segments. The methods range from simple cross-tabs to complex multivariate Accessible through unique analysis procedures. Examples include media/distribution (accessible) multiple discriminant analysis, logistic regression, correspondence analysis, Stable enough to exist over some and CHAID. planning horizon (stable)
WHICH SEGMENTS DO WE TARGET? Each of these six requirements must be met to justify a segmentation approach. These Once segments are identified and profiled, requirements are each discussed in turn. there is one more inescapable question: Do we target all of these segments or are our Size products suited for just a subset of the The segments must be large enough to segments? The answer to this question has sustain the additional cost incurred by many considerations: available resources, segmentation (like developing specialized current company image and its portability, communication messages). Segment size is
a relative question based on the size of the instance, there would be little trouble company, the size of the market, and the assigning individuals to the proper segment. buying power of each customer. Boeing, If however, a more sophisticated scheme is for example, would be able to define each used to identify segments, the researcher potential customer as a separate segment must be able to assign observations to and tailor their marketing mix to each segments independent of the basis variables. buyer's individual needs. A manufacturer Segments from post hoc assignment are of breakfast cereal, however, must segment formed on basis variables that are di cult the market on a more general basis. to measure--like attitudes or attribute importances. If the segments from a post Di erent hoc segmentation scheme cannot be The individual segments must be di erent profiled, the segments will be di cult to from each other. If all buyers in the market carry forward into future studies. behave basically the same, there is no need Accessible to segment the market. Likewise, if segments are found in which the behaviors across The segments must also be reachable. To groups are basically indistinguishable, the the extent that a segment exists, but cannot segmentation scheme is not accomplishing be educated about a product's potential what it should be. Markets for non- benefits by advertising, point of purchase, di erentiated, or commodity products or another method, it is of very little use to would not benefit from a segmented the company and of no more value than approach. A manufacturer of commodity the unsegmented market. products can more e ciently reach the entire market with one marketing strategy Stable than with several. There are, however, few Stability over time is also a reasonable markets that are absolutely not segmentable. concern for segmentation studies. The That is, almost every market has identifiable question is not so much that one person groups that behave di erently and by stays in the same segment all the time, but focusing on these di erences, companies that over time, the same basic segments can increase their marketing e ectiveness. exist, in about the same size, with the same Homogeneous basic profiles. Just as the segments need to be di erent, WHAT MAKES SEGMENTATION those potential buyers in the same segment RESEARCH SUCCESSFUL? need to be similar. That is, they must behave As pointed out above, almost any marketing similarly. This is necessary to treat the e ort can be made more e ective with segment as one entity. market segmentation, and almost any Actionable marketing research can be made more informative by analyzing the results at the Segments must be profiled so that potential segment level. buyers can be placed in a segment, based on easily measurable variables. This allows the While most people can clearly see the researcher to use the segmentation scheme benefits of segmentation, many are hesitant in future research without having to recreate to commit the time and e ort necessary to the segmentation scheme. These profiles perform a good segmentation study. Many could, for a priori segmentation schemes, people are tempted to perform a be the basis variables themselves. If rudimentary segmentation study hoping that segments were based on age and number the outcome will be as if the research design of children, or geographic region, for had been fully, and carefully, developed. This
fear has prompted some to refer to the “generic segmentation study, one done mindlessly and mechanically, following specific steps by rote under the assumption that clear-cut and insightful segments will somehow magically pop out of the welter of statistical output generated."1 The keys to successful segmentation research are:
Clearly defined goals and objectives
A substantial amount of up-front work
A thorough examination of segmentation dimensions
Well developed plans for follow- through activities (use what you have learned) Market segmentation is an extremely powerful strategic marketing tool. It must be treated as a strategic tool, though, and involve all levels of management and all future marketing decisions. When well considered, well planned and well executed, segmentation research can provide insights into your market that would otherwise go unnoticed and can provide increased e ciencies in all marketing e orts: product design, communication, distribution, and pricing. Market segmentation is a popular technique people likely to try this new technology. A in marketing, marketing research and limited segmentation scheme such as this is strategic planning. Market segmentation justified because the first people to try a new identifies unique groups of consumers in a product are rare, relatively homogeneous as market for various end-goals, including: early adopters/innovators, and do not require di erentiated marketing campaigns. Better understanding of consumers Further segmentation would be costly and Tailoring of marketing would require committing resources that communication messages could be more e ectively used elsewhere. New product development/line In this stage of the category life cycle further extensions/re-branding segmentation is not justified. Identifying potential economic/ As a product category matures, however, volumetric opportunities of new features are added to existing products, the segments the product gains distribution, more people The purpose of this document is to outline become aware of the product, and buyers when it is appropriate to consider segmentation become more savvy about the product and and the basic steps for executing its benefits. Likewise, competition increases. segmentation research. At this point, the seller must focus on more than selling a “new” product. The seller must WHEN TO CONSIDER SEGMENTATION also identify and meet the di ering needs of the market. The need for market segmentation can be shown by examining the category life cycle. As needs or desires expand into various stages In the early stages of a new product category, of heterogeneity, the seller is faced with there are few brands and few buyers. Market continuing to consider all buyers as similar or segmentation at this stage is simply separating adopt some type of market segmentation. those likely to try the new product from those The purpose of segmentation research, then, less likely. Media campaigns, advertising copy, is to find those unique subsets of the market and distribution channels are developed with that can be identified and treated as a group. this “likely to buy” group in mind. For example, The goal is to identify groups such that when the VCR was first introduced, the buying behavior within a group is similar marketing e orts were targeted at those (segment homogeneity) and the behavior
between groups is di erent. If the buyers have homogeneous needs, a company is best served by not segmenting the market. This is the case with most new (or innovative) products and some commodity products. Virtually every market, however, can benefit from a segmentation approach.
BASIC STEPS FOR which respondents are to be grouped based SEGMENTATION RESEARCH on age and household income. An a priori segmentation scheme could use the In deciding how to identify segments, the following categories shown above. researcher must answer four questions: In this scheme, each respondent could be What is the key objective of placed in the proper category as soon as the segmentation? these two pieces of information were known. What method should be used to segment the market? Post hoc segmentation, on the other hand, requires only that the variables on which the What variables, or basis, should be segmentation will be based (but not the used to segment the market? segment definitions) be identified. The How many segments should segments are then determined by using an be formed? analytical technique (such as cluster analysis) after all the data is collected. For example, if 1. Objective post hoc segments were to be formed on Frequently, the initial decision to conduct psychographics, the psychographic scores segmentation research is based on the for all respondents would be cluster analyzed. desire to better understand consumers. The segments would not be known until Though this is a key starting point, it is after running cluster analysis. This allows important to subsequently detail the groups to be formed based on "natural" strategic objectives to be met by this breaks in the data. This method is sometimes better understanding. This is the time to referred to as response-based. consider how the segments will be used The post hoc benefit segmentation approach once they have been identified. Taking the has become the preferred choice in most time to do this at the start of the process large “baseline” segmentation studies. By is essential for successful knowing what benefits buyers in a specific segmentation research. segment are looking to receive from a product, 2. Method the marketing mix is more clearly guided to appeal exactly to that target market. The method of segmentation refers to when the segments are defined. There are two 3. Basis methods of segmentation. After determining the Method of segmentation, They are a priori and post hoc. the researcher must determine what characteristics to use to identify the segments. Segmentation requires that respondents be These characteristics are referred to as the grouped based on some set of variables that basis of segmentation. Unlike the method are identified before data collection. In a question which had two possible answers, priori segmentation, not only are the the basis question is limited only by the variables identified, but so are the segment definitions. For example, consider a study in
researcher’s imagination. The following table illustrates some of the most frequently used bases of segmentation. Each of the major bases are discussed below:
Geographic Geographic segmentation calls for dividing the market into geographic regions. This allows the company to operate in only a few regions, because of limited demand or resources, or the company can tailor some element of the marketing mix by region. Food products are frequently varied by region. Either some products are only o ered in certain markets, or the di ering products are sold in the same container with the same advertising and promotion in di erent geographies.
Demographic Historically, the most common bases of segmentation have been demographic in nature – women between 18 and 49, for example. These schemes have frequently been motivated from a shopping or media behavior perspective and not one of brand or category choice behavior. While these schemes might adequately identify the frequent purchasers of a product category, they generally don’t fit the definition of segmentation as outlined above – dividing consumers by product choice behavior. Income, education, ethnicity, and occupation are all frequently used demographic segmentation variables.
Attitudinal Attitudinal variables are used based on the idea that opinions, traits, interests, and lifestyles influence buying behavior. Put another way, people with similar psychographic profiles should respond similarly to various marketing mixes. It is not clear that psychographics directly influence brand choice but are likely related to usage in several categories. Their application to market segmentation is well established.
Product Usage Product usage variables are also used to
MARKET SEGMENTATION | MARKETVISION RESEARCH
segment markets based on category usage, are derived based on causal factors relating to product usage, or brand loyalty. It is a the purchase process. People buy products, and common practice to segment the market decide between di erent brands, because of into users and non-users and also by level of the perceived unique benefits that one use (heavy vs. light). product can provide over another, and the relative importance of those benefits. Benefit Decision Process segmentation partitions the market into People decide which product to buy in segments based on benefits sought. The di erent ways. The influence that people result is that the people in one segment are exercise over decisions, as well as their brand seeking to realize some rather homogeneous loyalty, variety seeking behavior, information set of benefits with a particular product. search, distribution channel used and Those familiar with conjoint analysis can decision making units (centralized versus readily see the applicability for conjoint decentralized) all impact the brand that they to measure the relative importances will purchase. While this focus is more of the benefits sought. prevalent in business-to-business settings, certain elements are perfectly applicable The post hoc benefit segmentation approach to consumer segmentation as well. has become the preferred choice in most large “baseline” segmentation studies. By knowing Firmographic what benefits buyers in a specific segment The firmographic variables of SIC, revenue are looking to receive from a product, the and number of employees are standard bases marketing mix is more clearly guided to for industrial segmentation. Firmographic appeal exactly to that target market. segmentation is based on the idea that 4. Number of Segments companies can be targeted based simply on what industry they are in or how large Another question that must be answered in they are. This approach has been criticized any segmentation study is how many segments as largely ine ective. to use. The decision on how many segments has considerations from both statistical and Adoption Propensity managerial viewpoints. Segments can almost As shown when introducing the idea of always be made more homogeneous by segmentation, new product introductions increasing the number of segments. This could can likely be made more e cient by targeting continue until the number of segments becomes the early adopters of a product category. unmanageable. One practical guideline is Segmentation based on adoption propensity that unless the market can be broken into might be based on the propensity for the at least three groups, the advantages of specific product or category, or it might be segmenting are not cost e ective over a based on a buyer’s characteristic or propensity general market approach; and attempting to to adopt new product across several categories. market to more than seven segments is not This approach might even involve opinion practical for most companies. leadership – people whose opinions are sought There are obviously many possible ways to even though they might not be early adopters. segment a market. There is no single right Needs-Based or Benefit Segmentation answer and each situation can produce a di erent recommended “best” solution. What Needs-based or benefit segmentation is should be obvious is that although substantially di erent from the methods segmentation research involves substantial mentioned above. Instead of relying on scientific processes, it also relies inescapably descriptive factors (like region, age, income, on managerial judgment and the brand loyalty, or product usage), segments art of segmentation.
4
HOW TO USE MARKET SEGMENTS? competitive positioning, and market/industry dynamics. The basic approach, especially in To be most useful, the market segments just benefit segmentation, is to evaluate each identified should be tracked in all future segment’s desired product and evaluate research. To do this, researchers must develop whether current or new products can meet a way to make the segments identifiable and that segment’s needs without deteriorating meaningful so that they can be incorporated current franchise positioning. For example, in future e orts. This happens through many niche products which can command a segment profiles or descriptions. Segment price premium in their niche would require descriptions are useful for two primary reasons. price parity or even discounts when outside Most large segmentation studies of their niche. In this case, appealing to use variables that are not easily segments beyond the current niche is likely observable as the basis of unprofitable. This consideration of targeting segmentation. Measurements like must go beyond just the market’s desired large psychographic batteries, product specifications, but must also include part-worth utilities from conjoint and their willingness to consider a particular product preferences are frequently brand and distribution channels. used basis variables. Instead, the WHAT DOES SEGMENTATION segment profiles can be used to REQUIRE? identify the segments. E ective segmentation requires that truly Profiling the segments can provide di erent groups be identified. They must be: a validity check for the segmentation results themselves. The profiling Large enough to be economically stage can assist in the identification sustainable (size) of meaningless or artificial segments. If the profiles don't make sense Di erent enough to justify the from a managerial perspective, increased cost of segmentation the segmentation scheme must (di erent) be reconsidered. Similar enough within each segment Typical profiling variables are demographics, to be treated as one group psychographics, or other descriptive (homogeneous) variables. There are a number of proven Identifiable through segment research techniques that assist in profiling profiling (actionable) the segments. The methods range from simple cross-tabs to complex multivariate Accessible through unique analysis procedures. Examples include media/distribution (accessible) multiple discriminant analysis, logistic regression, correspondence analysis, Stable enough to exist over some and CHAID. planning horizon (stable)
WHICH SEGMENTS DO WE TARGET? Each of these six requirements must be met to justify a segmentation approach. These Once segments are identified and profiled, requirements are each discussed in turn. there is one more inescapable question: Do we target all of these segments or are our Size products suited for just a subset of the The segments must be large enough to segments? The answer to this question has sustain the additional cost incurred by many considerations: available resources, segmentation (like developing specialized current company image and its portability, communication messages). Segment size is
a relative question based on the size of the instance, there would be little trouble company, the size of the market, and the assigning individuals to the proper segment. buying power of each customer. Boeing, If however, a more sophisticated scheme is for example, would be able to define each used to identify segments, the researcher potential customer as a separate segment must be able to assign observations to and tailor their marketing mix to each segments independent of the basis variables. buyer's individual needs. A manufacturer Segments from post hoc assignment are of breakfast cereal, however, must segment formed on basis variables that are di cult the market on a more general basis. to measure--like attitudes or attribute importances. If the segments from a post Di erent hoc segmentation scheme cannot be The individual segments must be di erent profiled, the segments will be di cult to from each other. If all buyers in the market carry forward into future studies. behave basically the same, there is no need Accessible to segment the market. Likewise, if segments are found in which the behaviors across The segments must also be reachable. To groups are basically indistinguishable, the the extent that a segment exists, but cannot segmentation scheme is not accomplishing be educated about a product's potential what it should be. Markets for non- benefits by advertising, point of purchase, di erentiated, or commodity products or another method, it is of very little use to would not benefit from a segmented the company and of no more value than approach. A manufacturer of commodity the unsegmented market. products can more e ciently reach the entire market with one marketing strategy Stable than with several. There are, however, few Stability over time is also a reasonable markets that are absolutely not segmentable. concern for segmentation studies. The That is, almost every market has identifiable question is not so much that one person groups that behave di erently and by stays in the same segment all the time, but focusing on these di erences, companies that over time, the same basic segments can increase their marketing e ectiveness. exist, in about the same size, with the same Homogeneous basic profiles. Just as the segments need to be di erent, WHAT MAKES SEGMENTATION those potential buyers in the same segment RESEARCH SUCCESSFUL? need to be similar. That is, they must behave As pointed out above, almost any marketing similarly. This is necessary to treat the e ort can be made more e ective with segment as one entity. market segmentation, and almost any Actionable marketing research can be made more informative by analyzing the results at the Segments must be profiled so that potential segment level. buyers can be placed in a segment, based on easily measurable variables. This allows the While most people can clearly see the researcher to use the segmentation scheme benefits of segmentation, many are hesitant in future research without having to recreate to commit the time and e ort necessary to the segmentation scheme. These profiles perform a good segmentation study. Many could, for a priori segmentation schemes, people are tempted to perform a be the basis variables themselves. If rudimentary segmentation study hoping that segments were based on age and number the outcome will be as if the research design of children, or geographic region, for had been fully, and carefully, developed. This
fear has prompted some to refer to the “generic segmentation study, one done mindlessly and mechanically, following specific steps by rote under the assumption that clear-cut and insightful segments will somehow magically pop out of the welter of statistical output generated."1 The keys to successful segmentation research are:
Clearly defined goals and objectives
A substantial amount of up-front work
A thorough examination of segmentation dimensions
Well developed plans for follow- through activities (use what you have learned) Market segmentation is an extremely powerful strategic marketing tool. It must be treated as a strategic tool, though, and involve all levels of management and all future marketing decisions. When well considered, well planned and well executed, segmentation research can provide insights into your market that would otherwise go unnoticed and can provide increased e ciencies in all marketing e orts: product design, communication, distribution, and pricing. Market segmentation is a popular technique people likely to try this new technology. A in marketing, marketing research and limited segmentation scheme such as this is strategic planning. Market segmentation justified because the first people to try a new identifies unique groups of consumers in a product are rare, relatively homogeneous as market for various end-goals, including: early adopters/innovators, and do not require di erentiated marketing campaigns. Better understanding of consumers Further segmentation would be costly and Tailoring of marketing would require committing resources that communication messages could be more e ectively used elsewhere. New product development/line In this stage of the category life cycle further extensions/re-branding segmentation is not justified. Identifying potential economic/ As a product category matures, however, volumetric opportunities of new features are added to existing products, the segments the product gains distribution, more people The purpose of this document is to outline become aware of the product, and buyers when it is appropriate to consider segmentation become more savvy about the product and and the basic steps for executing its benefits. Likewise, competition increases. segmentation research. At this point, the seller must focus on more than selling a “new” product. The seller must WHEN TO CONSIDER SEGMENTATION also identify and meet the di ering needs of the market. The need for market segmentation can be shown by examining the category life cycle. As needs or desires expand into various stages In the early stages of a new product category, of heterogeneity, the seller is faced with there are few brands and few buyers. Market continuing to consider all buyers as similar or segmentation at this stage is simply separating adopt some type of market segmentation. those likely to try the new product from those The purpose of segmentation research, then, less likely. Media campaigns, advertising copy, is to find those unique subsets of the market and distribution channels are developed with that can be identified and treated as a group. this “likely to buy” group in mind. For example, The goal is to identify groups such that when the VCR was first introduced, the buying behavior within a group is similar marketing e orts were targeted at those (segment homogeneity) and the behavior
between groups is di erent. If the buyers have homogeneous needs, a company is best served by not segmenting the market. This is the case with most new (or innovative) products and some commodity products. Virtually every market, however, can benefit from a segmentation approach.
BASIC STEPS FOR which respondents are to be grouped based SEGMENTATION RESEARCH on age and household income. An a priori segmentation scheme could use the In deciding how to identify segments, the following categories shown above. researcher must answer four questions: In this scheme, each respondent could be What is the key objective of placed in the proper category as soon as the segmentation? these two pieces of information were known. What method should be used to segment the market? Post hoc segmentation, on the other hand, requires only that the variables on which the What variables, or basis, should be segmentation will be based (but not the used to segment the market? segment definitions) be identified. The How many segments should segments are then determined by using an be formed? analytical technique (such as cluster analysis) after all the data is collected. For example, if 1. Objective post hoc segments were to be formed on Frequently, the initial decision to conduct psychographics, the psychographic scores segmentation research is based on the for all respondents would be cluster analyzed. desire to better understand consumers. The segments would not be known until Though this is a key starting point, it is after running cluster analysis. This allows important to subsequently detail the groups to be formed based on "natural" strategic objectives to be met by this breaks in the data. This method is sometimes better understanding. This is the time to referred to as response-based. consider how the segments will be used The post hoc benefit segmentation approach once they have been identified. Taking the has become the preferred choice in most time to do this at the start of the process large “baseline” segmentation studies. By is essential for successful knowing what benefits buyers in a specific segmentation research. segment are looking to receive from a product, 2. Method the marketing mix is more clearly guided to appeal exactly to that target market. The method of segmentation refers to when the segments are defined. There are two 3. Basis methods of segmentation. After determining the Method of segmentation, They are a priori and post hoc. the researcher must determine what characteristics to use to identify the segments. Segmentation requires that respondents be These characteristics are referred to as the grouped based on some set of variables that basis of segmentation. Unlike the method are identified before data collection. In a question which had two possible answers, priori segmentation, not only are the the basis question is limited only by the variables identified, but so are the segment definitions. For example, consider a study in
researcher’s imagination. The following table illustrates some of the most frequently used bases of segmentation. Each of the major bases are discussed below:
Geographic Geographic segmentation calls for dividing the market into geographic regions. This allows the company to operate in only a few regions, because of limited demand or resources, or the company can tailor some element of the marketing mix by region. Food products are frequently varied by region. Either some products are only o ered in certain markets, or the di ering products are sold in the same container with the same advertising and promotion in di erent geographies.
Demographic Historically, the most common bases of segmentation have been demographic in nature – women between 18 and 49, for example. These schemes have frequently been motivated from a shopping or media behavior perspective and not one of brand or category choice behavior. While these schemes might adequately identify the frequent purchasers of a product category, they generally don’t fit the definition of segmentation as outlined above – dividing consumers by product choice behavior. Income, education, ethnicity, and occupation are all frequently used demographic segmentation variables.
Attitudinal Attitudinal variables are used based on the idea that opinions, traits, interests, and lifestyles influence buying behavior. Put another way, people with similar psychographic profiles should respond similarly to various marketing mixes. It is not clear that psychographics directly influence brand choice but are likely related to usage in several categories. Their application to market segmentation is well established.
Product Usage Product usage variables are also used to
segment markets based on category usage, are derived based on causal factors relating to product usage, or brand loyalty. It is a the purchase process. People buy products, and common practice to segment the market decide between di erent brands, because of into users and non-users and also by level of the perceived unique benefits that one use (heavy vs. light). product can provide over another, and the relative importance of those benefits. Benefit Decision Process segmentation partitions the market into People decide which product to buy in segments based on benefits sought. The di erent ways. The influence that people result is that the people in one segment are exercise over decisions, as well as their brand seeking to realize some rather homogeneous loyalty, variety seeking behavior, information set of benefits with a particular product. search, distribution channel used and Those familiar with conjoint analysis can decision making units (centralized versus readily see the applicability for conjoint decentralized) all impact the brand that they to measure the relative importances will purchase. While this focus is more of the benefits sought. prevalent in business-to-business settings, certain elements are perfectly applicable The post hoc benefit segmentation approach to consumer segmentation as well. has become the preferred choice in most large “baseline” segmentation studies. By knowing Firmographic what benefits buyers in a specific segment The firmographic variables of SIC, revenue are looking to receive from a product, the and number of employees are standard bases marketing mix is more clearly guided to for industrial segmentation. Firmographic appeal exactly to that target market. segmentation is based on the idea that 4. Number of Segments companies can be targeted based simply on what industry they are in or how large Another question that must be answered in they are. This approach has been criticized any segmentation study is how many segments as largely ine ective. to use. The decision on how many segments has considerations from both statistical and Adoption Propensity managerial viewpoints. Segments can almost As shown when introducing the idea of always be made more homogeneous by segmentation, new product introductions increasing the number of segments. This could can likely be made more e cient by targeting continue until the number of segments becomes the early adopters of a product category. unmanageable. One practical guideline is Segmentation based on adoption propensity that unless the market can be broken into might be based on the propensity for the at least three groups, the advantages of specific product or category, or it might be segmenting are not cost e ective over a based on a buyer’s characteristic or propensity general market approach; and attempting to to adopt new product across several categories. market to more than seven segments is not This approach might even involve opinion practical for most companies. leadership – people whose opinions are sought There are obviously many possible ways to even though they might not be early adopters. segment a market. There is no single right Needs-Based or Benefit Segmentation answer and each situation can produce a di erent recommended “best” solution. What Needs-based or benefit segmentation is should be obvious is that although substantially di erent from the methods segmentation research involves substantial mentioned above. Instead of relying on scientific processes, it also relies inescapably descriptive factors (like region, age, income, on managerial judgment and the brand loyalty, or product usage), segments art of segmentation.
MARKET SEGMENTATION | MARKETVISION RESEARCH
HOW TO USE MARKET SEGMENTS? competitive positioning, and market/industry dynamics. The basic approach, especially in To be most useful, the market segments just benefit segmentation, is to evaluate each identified should be tracked in all future segment’s desired product and evaluate research. To do this, researchers must develop whether current or new products can meet a way to make the segments identifiable and that segment’s needs without deteriorating meaningful so that they can be incorporated current franchise positioning. For example, in future e orts. This happens through many niche products which can command a segment profiles or descriptions. Segment price premium in their niche would require descriptions are useful for two primary reasons. price parity or even discounts when outside Most large segmentation studies of their niche. In this case, appealing to 1 use variables that are not easily segments beyond the current niche is likely observable as the basis of unprofitable. This consideration of targeting segmentation. Measurements like must go beyond just the market’s desired large psychographic batteries, product specifications, but must also include part-worth utilities from conjoint and their willingness to consider a particular product preferences are frequently brand and distribution channels. used basis variables. Instead, the WHAT DOES SEGMENTATION segment profiles can be used to REQUIRE? identify the segments. E ective segmentation requires that truly Profiling the segments can provide di erent groups be identified. They must be: 2 a validity check for the segmentation results themselves. The profiling Large enough to be economically stage can assist in the identification sustainable (size) of meaningless or artificial segments. If the profiles don't make sense Di erent enough to justify the from a managerial perspective, increased cost of segmentation the segmentation scheme must (di erent) be reconsidered. Similar enough within each segment Typical profiling variables are demographics, to be treated as one group psychographics, or other descriptive (homogeneous) variables. There are a number of proven Identifiable through segment research techniques that assist in profiling profiling (actionable) the segments. The methods range from simple cross-tabs to complex multivariate Accessible through unique analysis procedures. Examples include media/distribution (accessible) multiple discriminant analysis, logistic regression, correspondence analysis, Stable enough to exist over some and CHAID. planning horizon (stable)
WHICH SEGMENTS DO WE TARGET? Each of these six requirements must be met to justify a segmentation approach. These Once segments are identified and profiled, requirements are each discussed in turn. there is one more inescapable question: Do we target all of these segments or are our Size products suited for just a subset of the The segments must be large enough to segments? The answer to this question has sustain the additional cost incurred by many considerations: available resources, segmentation (like developing specialized current company image and its portability, communication messages). Segment size is
5
a relative question based on the size of the instance, there would be little trouble company, the size of the market, and the assigning individuals to the proper segment. buying power of each customer. Boeing, If however, a more sophisticated scheme is for example, would be able to define each used to identify segments, the researcher potential customer as a separate segment must be able to assign observations to and tailor their marketing mix to each segments independent of the basis variables. buyer's individual needs. A manufacturer Segments from post hoc assignment are of breakfast cereal, however, must segment formed on basis variables that are di cult the market on a more general basis. to measure--like attitudes or attribute importances. If the segments from a post Di erent hoc segmentation scheme cannot be The individual segments must be di erent profiled, the segments will be di cult to from each other. If all buyers in the market carry forward into future studies. behave basically the same, there is no need Accessible to segment the market. Likewise, if segments are found in which the behaviors across The segments must also be reachable. To groups are basically indistinguishable, the the extent that a segment exists, but cannot segmentation scheme is not accomplishing be educated about a product's potential what it should be. Markets for non- benefits by advertising, point of purchase, di erentiated, or commodity products or another method, it is of very little use to would not benefit from a segmented the company and of no more value than approach. A manufacturer of commodity the unsegmented market. products can more e ciently reach the entire market with one marketing strategy Stable than with several. There are, however, few Stability over time is also a reasonable markets that are absolutely not segmentable. concern for segmentation studies. The That is, almost every market has identifiable question is not so much that one person groups that behave di erently and by stays in the same segment all the time, but focusing on these di erences, companies that over time, the same basic segments can increase their marketing e ectiveness. exist, in about the same size, with the same Homogeneous basic profiles. Just as the segments need to be di erent, WHAT MAKES SEGMENTATION those potential buyers in the same segment RESEARCH SUCCESSFUL? need to be similar. That is, they must behave As pointed out above, almost any marketing similarly. This is necessary to treat the e ort can be made more e ective with segment as one entity. market segmentation, and almost any Actionable marketing research can be made more informative by analyzing the results at the Segments must be profiled so that potential segment level. buyers can be placed in a segment, based on easily measurable variables. This allows the While most people can clearly see the researcher to use the segmentation scheme benefits of segmentation, many are hesitant in future research without having to recreate to commit the time and e ort necessary to the segmentation scheme. These profiles perform a good segmentation study. Many could, for a priori segmentation schemes, people are tempted to perform a be the basis variables themselves. If rudimentary segmentation study hoping that segments were based on age and number the outcome will be as if the research design of children, or geographic region, for had been fully, and carefully, developed. This
fear has prompted some to refer to the “generic segmentation study, one done mindlessly and mechanically, following specific steps by rote under the assumption that clear-cut and insightful segments will somehow magically pop out of the welter of statistical output generated."1 The keys to successful segmentation research are:
Clearly defined goals and objectives
A substantial amount of up-front work
A thorough examination of segmentation dimensions
Well developed plans for follow- through activities (use what you have learned) Market segmentation is an extremely powerful strategic marketing tool. It must be treated as a strategic tool, though, and involve all levels of management and all future marketing decisions. When well considered, well planned and well executed, segmentation research can provide insights into your market that would otherwise go unnoticed and can provide increased e ciencies in all marketing e orts: product design, communication, distribution, and pricing. Market segmentation is a popular technique people likely to try this new technology. A in marketing, marketing research and limited segmentation scheme such as this is strategic planning. Market segmentation justified because the first people to try a new identifies unique groups of consumers in a product are rare, relatively homogeneous as market for various end-goals, including: early adopters/innovators, and do not require di erentiated marketing campaigns. Better understanding of consumers Further segmentation would be costly and Tailoring of marketing would require committing resources that communication messages could be more e ectively used elsewhere. New product development/line In this stage of the category life cycle further extensions/re-branding segmentation is not justified. Identifying potential economic/ As a product category matures, however, volumetric opportunities of new features are added to existing products, the segments the product gains distribution, more people The purpose of this document is to outline become aware of the product, and buyers when it is appropriate to consider segmentation become more savvy about the product and and the basic steps for executing its benefits. Likewise, competition increases. segmentation research. At this point, the seller must focus on more than selling a “new” product. The seller must WHEN TO CONSIDER SEGMENTATION also identify and meet the di ering needs of the market. The need for market segmentation can be shown by examining the category life cycle. As needs or desires expand into various stages In the early stages of a new product category, of heterogeneity, the seller is faced with there are few brands and few buyers. Market continuing to consider all buyers as similar or segmentation at this stage is simply separating adopt some type of market segmentation. those likely to try the new product from those The purpose of segmentation research, then, less likely. Media campaigns, advertising copy, is to find those unique subsets of the market and distribution channels are developed with that can be identified and treated as a group. this “likely to buy” group in mind. For example, The goal is to identify groups such that when the VCR was first introduced, the buying behavior within a group is similar marketing e orts were targeted at those (segment homogeneity) and the behavior
between groups is di erent. If the buyers have homogeneous needs, a company is best served by not segmenting the market. This is the case with most new (or innovative) products and some commodity products. Virtually every market, however, can benefit from a segmentation approach.
BASIC STEPS FOR which respondents are to be grouped based SEGMENTATION RESEARCH on age and household income. An a priori segmentation scheme could use the In deciding how to identify segments, the following categories shown above. researcher must answer four questions: In this scheme, each respondent could be What is the key objective of placed in the proper category as soon as the segmentation? these two pieces of information were known. What method should be used to segment the market? Post hoc segmentation, on the other hand, requires only that the variables on which the What variables, or basis, should be segmentation will be based (but not the used to segment the market? segment definitions) be identified. The How many segments should segments are then determined by using an be formed? analytical technique (such as cluster analysis) after all the data is collected. For example, if 1. Objective post hoc segments were to be formed on Frequently, the initial decision to conduct psychographics, the psychographic scores segmentation research is based on the for all respondents would be cluster analyzed. desire to better understand consumers. The segments would not be known until Though this is a key starting point, it is after running cluster analysis. This allows important to subsequently detail the groups to be formed based on "natural" strategic objectives to be met by this breaks in the data. This method is sometimes better understanding. This is the time to referred to as response-based. consider how the segments will be used The post hoc benefit segmentation approach once they have been identified. Taking the has become the preferred choice in most time to do this at the start of the process large “baseline” segmentation studies. By is essential for successful knowing what benefits buyers in a specific segmentation research. segment are looking to receive from a product, 2. Method the marketing mix is more clearly guided to appeal exactly to that target market. The method of segmentation refers to when the segments are defined. There are two 3. Basis methods of segmentation. After determining the Method of segmentation, They are a priori and post hoc. the researcher must determine what characteristics to use to identify the segments. Segmentation requires that respondents be These characteristics are referred to as the grouped based on some set of variables that basis of segmentation. Unlike the method are identified before data collection. In a question which had two possible answers, priori segmentation, not only are the the basis question is limited only by the variables identified, but so are the segment definitions. For example, consider a study in
researcher’s imagination. The following table illustrates some of the most frequently used bases of segmentation. Each of the major bases are discussed below:
Geographic Geographic segmentation calls for dividing the market into geographic regions. This allows the company to operate in only a few regions, because of limited demand or resources, or the company can tailor some element of the marketing mix by region. Food products are frequently varied by region. Either some products are only o ered in certain markets, or the di ering products are sold in the same container with the same advertising and promotion in di erent geographies.
Demographic Historically, the most common bases of segmentation have been demographic in nature – women between 18 and 49, for example. These schemes have frequently been motivated from a shopping or media behavior perspective and not one of brand or category choice behavior. While these schemes might adequately identify the frequent purchasers of a product category, they generally don’t fit the definition of segmentation as outlined above – dividing consumers by product choice behavior. Income, education, ethnicity, and occupation are all frequently used demographic segmentation variables.
Attitudinal Attitudinal variables are used based on the idea that opinions, traits, interests, and lifestyles influence buying behavior. Put another way, people with similar psychographic profiles should respond similarly to various marketing mixes. It is not clear that psychographics directly influence brand choice but are likely related to usage in several categories. Their application to market segmentation is well established.
Product Usage Product usage variables are also used to
segment markets based on category usage, are derived based on causal factors relating to product usage, or brand loyalty. It is a the purchase process. People buy products, and common practice to segment the market decide between di erent brands, because of into users and non-users and also by level of the perceived unique benefits that one use (heavy vs. light). product can provide over another, and the relative importance of those benefits. Benefit Decision Process segmentation partitions the market into People decide which product to buy in segments based on benefits sought. The di erent ways. The influence that people result is that the people in one segment are exercise over decisions, as well as their brand seeking to realize some rather homogeneous loyalty, variety seeking behavior, information set of benefits with a particular product. search, distribution channel used and Those familiar with conjoint analysis can decision making units (centralized versus readily see the applicability for conjoint decentralized) all impact the brand that they to measure the relative importances will purchase. While this focus is more of the benefits sought. prevalent in business-to-business settings, certain elements are perfectly applicable The post hoc benefit segmentation approach to consumer segmentation as well. has become the preferred choice in most large “baseline” segmentation studies. By knowing Firmographic what benefits buyers in a specific segment The firmographic variables of SIC, revenue are looking to receive from a product, the and number of employees are standard bases marketing mix is more clearly guided to for industrial segmentation. Firmographic appeal exactly to that target market. segmentation is based on the idea that 4. Number of Segments companies can be targeted based simply on what industry they are in or how large Another question that must be answered in they are. This approach has been criticized any segmentation study is how many segments as largely ine ective. to use. The decision on how many segments has considerations from both statistical and Adoption Propensity managerial viewpoints. Segments can almost As shown when introducing the idea of always be made more homogeneous by segmentation, new product introductions increasing the number of segments. This could can likely be made more e cient by targeting continue until the number of segments becomes the early adopters of a product category. unmanageable. One practical guideline is Segmentation based on adoption propensity that unless the market can be broken into might be based on the propensity for the at least three groups, the advantages of specific product or category, or it might be segmenting are not cost e ective over a based on a buyer’s characteristic or propensity general market approach; and attempting to to adopt new product across several categories. market to more than seven segments is not This approach might even involve opinion practical for most companies. leadership – people whose opinions are sought There are obviously many possible ways to even though they might not be early adopters. segment a market. There is no single right Needs-Based or Benefit Segmentation answer and each situation can produce a di erent recommended “best” solution. What Needs-based or benefit segmentation is should be obvious is that although substantially di erent from the methods segmentation research involves substantial mentioned above. Instead of relying on scientific processes, it also relies inescapably descriptive factors (like region, age, income, on managerial judgment and the brand loyalty, or product usage), segments art of segmentation.
HOW TO USE MARKET SEGMENTS? competitive positioning, and market/industry dynamics. The basic approach, especially in To be most useful, the market segments just benefit segmentation, is to evaluate each identified should be tracked in all future segment’s desired product and evaluate research. To do this, researchers must develop whether current or new products can meet a way to make the segments identifiable and that segment’s needs without deteriorating meaningful so that they can be incorporated current franchise positioning. For example, in future e orts. This happens through many niche products which can command a segment profiles or descriptions. Segment price premium in their niche would require descriptions are useful for two primary reasons. price parity or even discounts when outside Most large segmentation studies of their niche. In this case, appealing to use variables that are not easily segments beyond the current niche is likely observable as the basis of unprofitable. This consideration of targeting segmentation. Measurements like must go beyond just the market’s desired large psychographic batteries, product specifications, but must also include part-worth utilities from conjoint and their willingness to consider a particular product preferences are frequently brand and distribution channels. used basis variables. Instead, the WHAT DOES SEGMENTATION segment profiles can be used to REQUIRE? identify the segments. E ective segmentation requires that truly Profiling the segments can provide di erent groups be identified. They must be: a validity check for the segmentation results themselves. The profiling Large enough to be economically stage can assist in the identification sustainable (size) of meaningless or artificial segments. If the profiles don't make sense Di erent enough to justify the from a managerial perspective, increased cost of segmentation the segmentation scheme must (di erent) be reconsidered. Similar enough within each segment Typical profiling variables are demographics, to be treated as one group psychographics, or other descriptive (homogeneous) variables. There are a number of proven Identifiable through segment research techniques that assist in profiling profiling (actionable) the segments. The methods range from simple cross-tabs to complex multivariate Accessible through unique analysis procedures. Examples include media/distribution (accessible) multiple discriminant analysis, logistic regression, correspondence analysis, Stable enough to exist over some and CHAID. planning horizon (stable)
WHICH SEGMENTS DO WE TARGET? Each of these six requirements must be met to justify a segmentation approach. These Once segments are identified and profiled, requirements are each discussed in turn. there is one more inescapable question: Do we target all of these segments or are our Size products suited for just a subset of the The segments must be large enough to segments? The answer to this question has sustain the additional cost incurred by many considerations: available resources, segmentation (like developing specialized current company image and its portability, communication messages). Segment size is
MARKET SEGMENTATION | MARKETVISION RESEARCH
a relative question based on the size of the instance, there would be little trouble company, the size of the market, and the assigning individuals to the proper segment. buying power of each customer. Boeing, If however, a more sophisticated scheme is for example, would be able to define each used to identify segments, the researcher potential customer as a separate segment must be able to assign observations to and tailor their marketing mix to each segments independent of the basis variables. buyer's individual needs. A manufacturer Segments from post hoc assignment are of breakfast cereal, however, must segment formed on basis variables that are di cult the market on a more general basis. to measure--like attitudes or attribute importances. If the segments from a post Di erent hoc segmentation scheme cannot be The individual segments must be di erent profiled, the segments will be di cult to from each other. If all buyers in the market carry forward into future studies. behave basically the same, there is no need Accessible to segment the market. Likewise, if segments are found in which the behaviors across The segments must also be reachable. To groups are basically indistinguishable, the the extent that a segment exists, but cannot segmentation scheme is not accomplishing be educated about a product's potential what it should be. Markets for non- benefits by advertising, point of purchase, di erentiated, or commodity products or another method, it is of very little use to would not benefit from a segmented the company and of no more value than approach. A manufacturer of commodity the unsegmented market. products can more e ciently reach the entire market with one marketing strategy Stable than with several. There are, however, few Stability over time is also a reasonable markets that are absolutely not segmentable. concern for segmentation studies. The That is, almost every market has identifiable question is not so much that one person groups that behave di erently and by stays in the same segment all the time, but focusing on these di erences, companies that over time, the same basic segments can increase their marketing e ectiveness. exist, in about the same size, with the same Homogeneous basic profiles. Just as the segments need to be di erent, WHAT MAKES SEGMENTATION those potential buyers in the same segment RESEARCH SUCCESSFUL? need to be similar. That is, they must behave As pointed out above, almost any marketing similarly. This is necessary to treat the e ort can be made more e ective with segment as one entity. market segmentation, and almost any Actionable marketing research can be made more informative by analyzing the results at the Segments must be profiled so that potential segment level. buyers can be placed in a segment, based on easily measurable variables. This allows the While most people can clearly see the researcher to use the segmentation scheme benefits of segmentation, many are hesitant in future research without having to recreate to commit the time and e ort necessary to the segmentation scheme. These profiles perform a good segmentation study. Many could, for a priori segmentation schemes, people are tempted to perform a be the basis variables themselves. If rudimentary segmentation study hoping that segments were based on age and number the outcome will be as if the research design of children, or geographic region, for had been fully, and carefully, developed. This
6
fear has prompted some to refer to the “generic segmentation study, one done mindlessly and mechanically, following specific steps by rote under the assumption that clear-cut and insightful segments will somehow magically pop out of the welter of statistical output generated."1 The keys to successful segmentation research are:
Clearly defined goals and objectives
A substantial amount of up-front work
A thorough examination of segmentation dimensions
Well developed plans for follow- through activities (use what you have learned) Market segmentation is an extremely powerful strategic marketing tool. It must be treated as a strategic tool, though, and involve all levels of management and all future marketing decisions. When well considered, well planned and well executed, segmentation research can provide insights into your market that would otherwise go unnoticed and can provide increased e ciencies in all marketing e orts: product design, communication, distribution, and pricing. Market segmentation is a popular technique people likely to try this new technology. A in marketing, marketing research and limited segmentation scheme such as this is strategic planning. Market segmentation justified because the first people to try a new identifies unique groups of consumers in a product are rare, relatively homogeneous as market for various end-goals, including: early adopters/innovators, and do not require di erentiated marketing campaigns. Better understanding of consumers Further segmentation would be costly and Tailoring of marketing would require committing resources that communication messages could be more e ectively used elsewhere. New product development/line In this stage of the category life cycle further extensions/re-branding segmentation is not justified. Identifying potential economic/ As a product category matures, however, volumetric opportunities of new features are added to existing products, the segments the product gains distribution, more people The purpose of this document is to outline become aware of the product, and buyers when it is appropriate to consider segmentation become more savvy about the product and and the basic steps for executing its benefits. Likewise, competition increases. segmentation research. At this point, the seller must focus on more than selling a “new” product. The seller must WHEN TO CONSIDER SEGMENTATION also identify and meet the di ering needs of the market. The need for market segmentation can be shown by examining the category life cycle. As needs or desires expand into various stages In the early stages of a new product category, of heterogeneity, the seller is faced with there are few brands and few buyers. Market continuing to consider all buyers as similar or segmentation at this stage is simply separating adopt some type of market segmentation. those likely to try the new product from those The purpose of segmentation research, then, less likely. Media campaigns, advertising copy, is to find those unique subsets of the market and distribution channels are developed with that can be identified and treated as a group. this “likely to buy” group in mind. For example, The goal is to identify groups such that when the VCR was first introduced, the buying behavior within a group is similar marketing e orts were targeted at those (segment homogeneity) and the behavior
between groups is di erent. If the buyers have homogeneous needs, a company is best served by not segmenting the market. This is the case with most new (or innovative) products and some commodity products. Virtually every market, however, can benefit from a segmentation approach.
BASIC STEPS FOR which respondents are to be grouped based SEGMENTATION RESEARCH on age and household income. An a priori segmentation scheme could use the In deciding how to identify segments, the following categories shown above. researcher must answer four questions: In this scheme, each respondent could be What is the key objective of placed in the proper category as soon as the segmentation? these two pieces of information were known. What method should be used to segment the market? Post hoc segmentation, on the other hand, requires only that the variables on which the What variables, or basis, should be segmentation will be based (but not the used to segment the market? segment definitions) be identified. The How many segments should segments are then determined by using an be formed? analytical technique (such as cluster analysis) after all the data is collected. For example, if 1. Objective post hoc segments were to be formed on Frequently, the initial decision to conduct psychographics, the psychographic scores segmentation research is based on the for all respondents would be cluster analyzed. desire to better understand consumers. The segments would not be known until Though this is a key starting point, it is after running cluster analysis. This allows important to subsequently detail the groups to be formed based on "natural" strategic objectives to be met by this breaks in the data. This method is sometimes better understanding. This is the time to referred to as response-based. consider how the segments will be used The post hoc benefit segmentation approach once they have been identified. Taking the has become the preferred choice in most time to do this at the start of the process large “baseline” segmentation studies. By is essential for successful knowing what benefits buyers in a specific segmentation research. segment are looking to receive from a product, 2. Method the marketing mix is more clearly guided to appeal exactly to that target market. The method of segmentation refers to when the segments are defined. There are two 3. Basis methods of segmentation. After determining the Method of segmentation, They are a priori and post hoc. the researcher must determine what characteristics to use to identify the segments. Segmentation requires that respondents be These characteristics are referred to as the grouped based on some set of variables that basis of segmentation. Unlike the method are identified before data collection. In a question which had two possible answers, priori segmentation, not only are the the basis question is limited only by the variables identified, but so are the segment definitions. For example, consider a study in
researcher’s imagination. The following table illustrates some of the most frequently used bases of segmentation. Each of the major bases are discussed below:
Geographic Geographic segmentation calls for dividing the market into geographic regions. This allows the company to operate in only a few regions, because of limited demand or resources, or the company can tailor some element of the marketing mix by region. Food products are frequently varied by region. Either some products are only o ered in certain markets, or the di ering products are sold in the same container with the same advertising and promotion in di erent geographies.
Demographic Historically, the most common bases of segmentation have been demographic in nature – women between 18 and 49, for example. These schemes have frequently been motivated from a shopping or media behavior perspective and not one of brand or category choice behavior. While these schemes might adequately identify the frequent purchasers of a product category, they generally don’t fit the definition of segmentation as outlined above – dividing consumers by product choice behavior. Income, education, ethnicity, and occupation are all frequently used demographic segmentation variables.
Attitudinal Attitudinal variables are used based on the idea that opinions, traits, interests, and lifestyles influence buying behavior. Put another way, people with similar psychographic profiles should respond similarly to various marketing mixes. It is not clear that psychographics directly influence brand choice but are likely related to usage in several categories. Their application to market segmentation is well established.
Product Usage Product usage variables are also used to
segment markets based on category usage, are derived based on causal factors relating to product usage, or brand loyalty. It is a the purchase process. People buy products, and common practice to segment the market decide between di erent brands, because of into users and non-users and also by level of the perceived unique benefits that one use (heavy vs. light). product can provide over another, and the relative importance of those benefits. Benefit Decision Process segmentation partitions the market into People decide which product to buy in segments based on benefits sought. The di erent ways. The influence that people result is that the people in one segment are exercise over decisions, as well as their brand seeking to realize some rather homogeneous loyalty, variety seeking behavior, information set of benefits with a particular product. search, distribution channel used and Those familiar with conjoint analysis can decision making units (centralized versus readily see the applicability for conjoint decentralized) all impact the brand that they to measure the relative importances will purchase. While this focus is more of the benefits sought. prevalent in business-to-business settings, certain elements are perfectly applicable The post hoc benefit segmentation approach to consumer segmentation as well. has become the preferred choice in most large “baseline” segmentation studies. By knowing Firmographic what benefits buyers in a specific segment The firmographic variables of SIC, revenue are looking to receive from a product, the and number of employees are standard bases marketing mix is more clearly guided to for industrial segmentation. Firmographic appeal exactly to that target market. segmentation is based on the idea that 4. Number of Segments companies can be targeted based simply on what industry they are in or how large Another question that must be answered in they are. This approach has been criticized any segmentation study is how many segments as largely ine ective. to use. The decision on how many segments has considerations from both statistical and Adoption Propensity managerial viewpoints. Segments can almost As shown when introducing the idea of always be made more homogeneous by segmentation, new product introductions increasing the number of segments. This could can likely be made more e cient by targeting continue until the number of segments becomes the early adopters of a product category. unmanageable. One practical guideline is Segmentation based on adoption propensity that unless the market can be broken into might be based on the propensity for the at least three groups, the advantages of specific product or category, or it might be segmenting are not cost e ective over a based on a buyer’s characteristic or propensity general market approach; and attempting to to adopt new product across several categories. market to more than seven segments is not This approach might even involve opinion practical for most companies. leadership – people whose opinions are sought There are obviously many possible ways to even though they might not be early adopters. segment a market. There is no single right Needs-Based or Benefit Segmentation answer and each situation can produce a di erent recommended “best” solution. What Needs-based or benefit segmentation is should be obvious is that although substantially di erent from the methods segmentation research involves substantial mentioned above. Instead of relying on scientific processes, it also relies inescapably descriptive factors (like region, age, income, on managerial judgment and the brand loyalty, or product usage), segments art of segmentation.
HOW TO USE MARKET SEGMENTS? competitive positioning, and market/industry dynamics. The basic approach, especially in To be most useful, the market segments just benefit segmentation, is to evaluate each identified should be tracked in all future segment’s desired product and evaluate research. To do this, researchers must develop whether current or new products can meet a way to make the segments identifiable and that segment’s needs without deteriorating meaningful so that they can be incorporated current franchise positioning. For example, in future e orts. This happens through many niche products which can command a segment profiles or descriptions. Segment price premium in their niche would require descriptions are useful for two primary reasons. price parity or even discounts when outside Most large segmentation studies of their niche. In this case, appealing to use variables that are not easily segments beyond the current niche is likely observable as the basis of unprofitable. This consideration of targeting segmentation. Measurements like must go beyond just the market’s desired large psychographic batteries, product specifications, but must also include part-worth utilities from conjoint and their willingness to consider a particular product preferences are frequently brand and distribution channels. used basis variables. Instead, the WHAT DOES SEGMENTATION segment profiles can be used to REQUIRE? identify the segments. E ective segmentation requires that truly Profiling the segments can provide di erent groups be identified. They must be: a validity check for the segmentation results themselves. The profiling Large enough to be economically stage can assist in the identification sustainable (size) of meaningless or artificial segments. If the profiles don't make sense Di erent enough to justify the from a managerial perspective, increased cost of segmentation the segmentation scheme must (di erent) be reconsidered. Similar enough within each segment Typical profiling variables are demographics, to be treated as one group psychographics, or other descriptive (homogeneous) variables. There are a number of proven Identifiable through segment research techniques that assist in profiling profiling (actionable) the segments. The methods range from simple cross-tabs to complex multivariate Accessible through unique analysis procedures. Examples include media/distribution (accessible) multiple discriminant analysis, logistic regression, correspondence analysis, Stable enough to exist over some and CHAID. planning horizon (stable)
WHICH SEGMENTS DO WE TARGET? Each of these six requirements must be met to justify a segmentation approach. These Once segments are identified and profiled, requirements are each discussed in turn. there is one more inescapable question: Do we target all of these segments or are our Size products suited for just a subset of the The segments must be large enough to segments? The answer to this question has sustain the additional cost incurred by many considerations: available resources, segmentation (like developing specialized current company image and its portability, communication messages). Segment size is
a relative question based on the size of the instance, there would be little trouble company, the size of the market, and the assigning individuals to the proper segment. buying power of each customer. Boeing, If however, a more sophisticated scheme is for example, would be able to define each used to identify segments, the researcher potential customer as a separate segment must be able to assign observations to and tailor their marketing mix to each segments independent of the basis variables. buyer's individual needs. A manufacturer Segments from post hoc assignment are of breakfast cereal, however, must segment formed on basis variables that are di cult the market on a more general basis. to measure--like attitudes or attribute importances. If the segments from a post Di erent hoc segmentation scheme cannot be The individual segments must be di erent profiled, the segments will be di cult to from each other. If all buyers in the market carry forward into future studies. behave basically the same, there is no need Accessible to segment the market. Likewise, if segments are found in which the behaviors across The segments must also be reachable. To groups are basically indistinguishable, the the extent that a segment exists, but cannot segmentation scheme is not accomplishing be educated about a product's potential what it should be. Markets for non- benefits by advertising, point of purchase, di erentiated, or commodity products or another method, it is of very little use to would not benefit from a segmented the company and of no more value than approach. A manufacturer of commodity the unsegmented market. products can more e ciently reach the entire market with one marketing strategy Stable than with several. There are, however, few Stability over time is also a reasonable markets that are absolutely not segmentable. concern for segmentation studies. The That is, almost every market has identifiable question is not so much that one person groups that behave di erently and by stays in the same segment all the time, but focusing on these di erences, companies that over time, the same basic segments can increase their marketing e ectiveness. exist, in about the same size, with the same Homogeneous basic profiles. Just as the segments need to be di erent, WHAT MAKES SEGMENTATION those potential buyers in the same segment RESEARCH SUCCESSFUL? need to be similar. That is, they must behave As pointed out above, almost any marketing similarly. This is necessary to treat the e ort can be made more e ective with segment as one entity. market segmentation, and almost any Actionable marketing research can be made more informative by analyzing the results at the Segments must be profiled so that potential segment level. buyers can be placed in a segment, based on easily measurable variables. This allows the While most people can clearly see the researcher to use the segmentation scheme benefits of segmentation, many are hesitant in future research without having to recreate to commit the time and e ort necessary to the segmentation scheme. These profiles perform a good segmentation study. Many could, for a priori segmentation schemes, people are tempted to perform a be the basis variables themselves. If rudimentary segmentation study hoping that segments were based on age and number the outcome will be as if the research design of children, or geographic region, for had been fully, and carefully, developed. This
MARKET SEGMENTATION | MARKETVISION RESEARCH
fear has prompted some to refer to the “generic segmentation study, one done mindlessly and mechanically, following specific steps by rote under the assumption that clear-cut and insightful segments will somehow magically pop out of the welter of statistical output generated."1 The keys to successful segmentation research are:
1 Clearly defined goals and objectives
2 A substantial amount of up-front work A thorough examination of 3 segmentation dimensions
Well developed plans for follow- 4 through activities (use what you have learned) Market segmentation is an extremely powerful strategic marketing tool. It must be treated as a strategic tool, though, and involve all levels of management and all future marketing decisions. When well considered, well planned and well executed, segmentation research can provide insights into your market that would otherwise go unnoticed and can provide increased e ciencies in all marketing e orts: product design, communication, distribution, and pricing.
1 Haley, Russell. (1984), "Benefit Segmentation: Backwards and Forwards," Journal of Advertising Research, 24 (February/March), 19-25.
7