May 21, 2013
BrainTrust Query: What Can Customer Segmentation Not Do?
Through a special arrangement, what follows is an excerpt of an article from WayfinD, a quarterly e-magazine filled with insights, trends and predictions from the retail and foodservice experts at WD Partners.
Too much information, running through my brain.
Too much information, driving me insane.
— The Police
Segmentation can help a retailer or brand make better marketing and design decisions by better understanding what makes their customers tick and sizing up opportunity segments. But it’s important to know, going in to a segmentation program, what you can realistically expect and accomplish.
Here are four objectives segmentation can’t deliver:
Creating a monolithic target segment: Unless you offer extremely niche products or services, your segmentation will be a bit messy. The most effective segments typically divide the total addressable market (TAM) into 6-10 segments, and then highlight 3-4 of those as the target segments. A segmentation helps organizations makes choices — it does not dictate decisions.
Defining your brand strategy: Your brand strategy needs to be the core of your organization. It should define everything you do, from product development to SKU rationalization to ad campaigns. The goal of segmentation is to highlight differences between groups. Your brand strategy needs to be the consistent idea that unites and spans across segments and geographies. Your segmentation should be one input when creating a relevant and differentiated brand strategy. Creating a modular brand strategy for different segments and geographies is a potential recipe for disaster.
Determining innovation: Segmentation is a snapshot in time that can help drive innovation by highlighting the key attitudes and behaviors of your priority segments. Creating innovation in products and customer experiences requires an understanding of your audience coupled with future thinking that takes into account a variety of inputs — both tangible and intangible.
Replacing ongoing consumer/shopper research: Segmentation is typically done when an organization is at a pivot point, whether it’s a new store design or a comprehensive rebranding effort. This holistic assessment of your customers and your business strategies helps your organization make big decisions. As a rule of thumb, a segmentation remains relevant for 3-5 years and then needs to be refreshed to account for shifts in your customers’ lifestyles and market fluctuations. In addition to segmentation, an organization needs to consistently collect data (in various forms) to make sure your brand, your product, and your stores are aligning with the ever-evolving needs and wants of your customers.
Like any tool, consumer segmentation needs to be used appropriately and in the right situations. Any information, including segmentation, needs to be used judiciously with ample amounts of common sense.
BrainTrust
Recent Discussions
Discussion Questions
What can customer segmentation do and not do? What irrational expectations around segmentation would you add to the list above?
Poll
Segmentation is great when (a) segments really exist that are not just a figment of someone’s statistical imagination and (b) it produces groups that are truly different in their attitudes, behaviors, shopping patterns, targetability, and so forth. Nothing says that everything can be segmented in a rational and useful way.
Marketers seem to have an addictive desire to segment, a bit ironic when you consider that consumer purchasing behavior is getting less and less predictable and consumers themselves are falling into multiple segments, often in the space of a single day.
Over ten years ago in our book “The Myth of Excellence,” Fred Crawford and I described the emergence of what we termed the “Instavidual”—people who voluntarily moved from traditional segmentation point to traditional segmentation point. Compound this “instavidualism” with the emergence of multiple platforms, channels and devices and segmentation becomes exponentially more difficult.
Finally, add the fact that lifestyles are changing faster than marketing and you see a generational, ethnic and class disconnect further muddling the segmentation waters.
So, what segmentation can’t do is get to the truth of an individual shopper as a person and—by extension—this means it isn’t going to be too much more successful when it aggregates those individual shoppers.
The only way to get to the truth of the individual is to communicate with him or her, so we should be focusing more on devising better feedback loops and less on which box we can force fit our customers into.
As companies struggle to manage data, it’s good business practice to employ features of segmentation to ascertain and glean insights. Segmentation allows for companies to find a common thread and differentiate information. It’s logical and practical and you can start to understand and pull real truths from it.
But Tyler Anderson is right: it cannot do your work for you. If you don’t understand your market, your segmentation insights won’t mean much. Without a clear brand strategy, what are you segmenting, and why? Segmentation can find commonalities to spearhead innovation, but not create advances. Segmentation is an instrument, along with many other mechanisms, to pool data points and make sense of an ever changing marketplace.
Segmentation won’t take the place of the points outlined in the article, but it definitely helps to make sense of common business questions and finding answers to those problems.
Segmentation is sometimes thought of as nirvana. If I do the segmentation I’ll have the answer. But you won’t have an answer, if you don’t know the question you are asking.
Segmentation research should be designed to uncover growth opportunities. It should be forward thinking not a retrospective on the way life used to be or might be currently. Effective segmentation research requires an achievable objective, a well thought out design and very importantly creative analytics dedicated to the overarching mission of the company.
If the goal is simply to understand existing customers better, the segmentation should allow the retailer to determine what if anything differentiates their high revenue customers from those that spend less money in their stores. That creates the opportunity to develop strategies to generate more revenue from big spenders and small spenders alike.
However, the retailer may have a grander goal. For instance the marketer may want to understand the competitive landscape and segment by consumers who shop in one retail brand or another (or more frequently). That creates the opportunity to generate loyalty among current shoppers and develop strategies to attract new customers.
Bottom line, without a purpose that can drive growth, the segmentation is a waste of time and money.
Segmentation allows retailers and brands to put consumers into convenient buckets for advertising/marketing purposes. The better the definition of each segment, the more tactics can be refined to reach it.
Segmentation needs to go beyond age, sex and geography. It needs to dive into relevancy, usage patterns and frequency of purchase.
Segments need to constantly be reevaluated. Waiting 3-5 years is too long…a lifetime in the digital age.
The ultimate goal for marketers should be to deliver a customized message to each consumer.
You can segment yourself to death these days. I know from work within and outside the supermarket industry that more retailers are dealing with micro-segmentation—groups within the groups.
Retailers should be paying closer attention to the data available to them these days and segmentation is not something you do once for a new store then forget. It needs to be constant and consistent, and I would suggest that every 3-5 years is not enough.
But let’s face facts. Segmentation doesn’t keep the shelves stocked, the store clean, the employees friendly and helpful, or shorten the lines at the checkout.
Marketing quests today still aren’t moving as quickly as the consumers are changing. That can cause an informational imbalance. In dealing with customer segmentation, may we remember that no one has the same fingerprints. True, even as more revealing data is pooled, the ultimate decision by a consumer is still the consumer’s … and thus customer segmentation needs to be used judiciously.
In today’s retail environment, it is clearly not enough to segment. I like the terms used by service designers, ‘roles’ and ‘players’. Every brand needs to understand what the roles and who the players are that come in contact with their brands.
This calls for more than demographic information. It also calls for an understanding of how various consumers interact with the brand. What is the profile in each role? Keep in mind that a single individual may take on different roles dependent on their need at that particular time.
Once the retailer can map this out, the information is golden in how each channel is constructed, merchandised and staffed. IMHO
Just wanted to elaborate a bit re: WayfinD’s contribution here, which I found good, but missing in some key areas:
1. You do not necessarily need to offer niche products and services to avoid monolithic segmentation modeling results. Remember, the segmentation effort is not about the brand or the retailer, not about brands that are offered, it’s about your customers. Monoliths like Hellmann’s, Heinz, most retailers such as Macy’s, Nordstrom’s, Neiman Marcus, can and will engender meaningful, highly differentiable segments of their customer base only if their input bases for segmentation consider and contain many different kinds of input variables: attitudinal, emotional need states, psychographics, lifestyles, shopper styles and behavioral (e.g., situational, ‘moments’) variables. Remember: garbage in, garbage out. You must do the due diligence at the a priori stages (qualitative, planning with stakeholders) prior to the quantitative data gathering phase, otherwise, you will end up with a meaningless effort if your input going in is non-differentiable.
2. Individual resulting segment strategies are perfectly acceptable, given that the common denominator across target segments is the brand’s overall equity, and that this equity be kept intact throughout all strategic development for the individual segments to be targeted. Usually, the strategy is to maintain and enhance each key segment’s equity perceptions, while optimizing each one’s (emotional) brand loyalty, customer satisfaction and retention–all the while maintaining the essence of equity, and all the while attempting to expand core customer usage. The overall goal with each important target segment is to minimize alienation and defection, expand usage and grow the number of customers overall. And yes, in some cases, different strategies aimed at each valuable target segment can be required, as long as each segment strategy is not incongruous with or off overall brand strategy.
3. Mention about the important contribution made by segmentation research to resource allocation is infrequent. Segments that account for high-density, high-index dollar volume spending and consumption in relation to their actual incidence in the brand’s space are considered more important than others. Those that are considered less or non-important will be deemed to warrant less in the way of advertising, marketing and innovation efforts. Segmentation is an important tool to help guide your allocation of dollars.
4. Let the segmentation results and ensuing storylines guide subsequent research (new product concept screening and evaluation, trackers, IHUT product testing, ad communication studies, new packaging and logo design, spokesperson selection, etc.). Make sure that key segment players’ attitudes and responses are represented in such studies by use of the segment model algorithm for study screening and recruitment purposes.
5. Make sure to get corporate stakeholders to buy into the study at early phases, by having them participate and contribute to inputting into the info-seek process. Have them provide their ‘wish-list’ objectives, insights, hypotheses. Where, at the end of the study, do they want to be with this? Having them provide this end-study wish list enables you to track back to insure that this end-goal is reached.
To me, customer segmentation has lost half its value. We used to use segmentation for research and insight, and also to create unique experiences/solutions for each segment.
Segmentation is still a very useful tool for research, and for getting insight out of big data pools, but the days of taking action on specific customer segments is at an end. Technology now lets us take action on individuals rather than aggregated segments of shoppers.
I’m generally appreciative of any discussion that questions conventional wisdom. This thread is a good example.
Segmentation seeks to identify enduring differences between target-able groups. In the early days this was based principally on geo-demographics, later on geo-demo plus psychographics, more recently on behavioral metrics. Some targeting is generally better than no targeting, so these activities have harvested low-hanging fruit for marketers.
Any segmentation, however sophisticated, is at best a snapshot in time. One of the stupendous lessons we must take from SoMoLoMe and Big Data is that consumer behavior reality speeds up and changes much faster than our decision cycles.
So any brand mission statement that begins, “Our customer is…” should be considered suspect if it’s more than a minute old. Three to five years is delusional.
Maybe segmentation needs to give way to snapshot analysis and data mining. That is, periodically look at all available descriptive data about who is buying our product now. Then search the universe of shoppers for others who match those traits and target them.
Ultimately only three segments matter then: (1) The folks who bought your brand this week. (2) Other people who resemble those folks, but who did not buy your brand. (3) Everybody else.
Tyler Anderson raises some very interesting functions and limitations of customer segmentation. Ultimately the strength of any segment lies in how tightly homogenous the segment actually is so that common treatment of each segment member results in relevance and enhanced shopper engagement.
Retailers that still use RFM (Recency, Frequency, Monetary Value) segmentation likely already know that there is much variation in terms of lifestyles, priorities, and product preferences within RFM segments as these segments are devised with just one or two variables. I would encourage any retailer who is still robustly relying on a RFM approach to look at other key determinants of shopping behavior, optimally derived a blend of customer transactional data mining and qualitative research designed to weigh the impact of each determinant on actual shopping behavior.
Conversely, the worst of all segmentation practices results in having shopper segments that are so esoteric and specific that they do not practically lend themselves to action that will benefit the business. I have seen both “polar” extremes in play and neither get the job done.
The online world poses a new challenge: personalization, which in principle works totally reciprocally to segmentation. In segmentation you find similarities among groups of people where as in personalization you are focusing on each person as an individual. However, while creating offers or recommendations, especially in the online world, the struggle would be to draw that line to make it profitable and meaningful for both the retailer and the shopper.
A good segmentation is the one that understands the differences among segments and the reason behind it for the specific event/offer/any merchandising decision. What does not work is the cookie cutter or stereotypical approach to segmentation. Segmentation should appreciate the situational differences; not every individual would respond to a trigger in the same way.
Segmentation can be a daunting task. It can also dilute your brand. Smart brands create a consistent message, even if they are marketing through different channels. They may change the visual aspect, and even some words of an advert, but will still maintain their core message.
Segmentation work is needed in certain industries. For example, a grocery store chain may stock certain items based on the local demographic. A hardware store probably doesn’t stock snow shovels in Miami, Florida, but snow shovels are a staple in Montana. But what probably doesn’t change with the grocery store or the hardware store is their core culture.
For example, if Ace Hardware stores want to be known as the Helpful Hardware place, they are going to helpful in every store, regardless of the demographics.
Put simply, customer segmentation is key to effective promotional planning and price optimization. When you try to create effective pricing for customers as a whole, you’re mixing people who are price sensitive with those who aren’t—and their reactions could be dramatically different to a given promotion. Yet when you measure price elasticity at the customer segment level, not just the product level, your forecast accuracy is much greater and as noted above, you can make strategic choices that provide the most value within your customers segments. This ultimately builds long-term shopper loyalty.
Some great comments below have been made on traditional views of segmentation. I believe the perspective we have talked about for years, that is “A segment of one” is now within our reach. Targeting, marketing, advertising and selling to individual shoppers, as they evolve, and being nimble enough to evolve with them… AND anticipate their evolution, is the new reality with tools available today.
Segmentation is a great marketing tool for both traditional and digital. It is not a silver bullet and does not replace common sense. Segmentation is not everything; it is not a brand strategy nor does it replace innovation or other research. To an experienced marketer, these things seem to be self-evident, however, I suppose they bear saying. Segmentation does provide a solid tool for marketers to be able to understand and communicate with their customers. In the digital age customization is cheap, easy and most importantly expected by customers, segmentation is not only recommended, but almost a requirement.
The idea that a single segmentation base can be usable input for different decisions, i.e. if you are concerned with making pricing decisions, segment on price/sensitivity bases; if a media decision, use listening/viewing/reading, etc. bases. If they’re competitive positioning decisions, another set. Etc.
Customer segmentation is an excellent tool for understanding customers and eventually targeting them. Key is to know for what purpose are you doing customer segmentation. Is it for introducing new products or is it for doing specific campaigns? The way that you do segmentation will change. A unique customer may belong to multiple segments as per the need. Also, as rightly pointed out, segmentation is not a one-time exercise—it changes all the time. What was true 3 years ago may not be close to truth today.
So here’s my 2 cents : (1) Make segmentation relevant (2) Do it every year (3) Couple it with other analysis (e.g. market structure analysis).
The most common mistake I see with segmentation is that it is not actionable. In search of more clever models, we hunt for significant factors and lose sight of the customer.
In marketing, we find the variables that best discriminate are the ones that are most difficult to target, such as attitudes. But the factors easiest to target, like demographics and firmographics, are applied to media buys because that’s where large audiences can be found.
In selling, we borrow from marketing surveys rather than purchase patterns. But the best targets are not those who rate your value prop highly but those who exhibit desired purchase behavior such as R-F-M.
In product development, we aggregate singular preferences into abstract clusters and lose the clarity of individuals’ problems. But the best solution to stale product development is to spend more time with individual buyers
The antidote is to refrain from narrow appeals that come from thin-slicing the data.
More Discussions
Is Meta Committing to Retail via Meta Lab Stores, or Are Other Goals More Important?
Is Meta playing in retail, actually committing to retail, or looking for more data?Aerie Recruits Pamela Anderson To Highlight No-AI Models Promise
American Eagle’s Aerie banner has recruited former “Baywatch” star Pamela Anderson to broadcast the lifestyle brand’s…How Timely Is the Planned Lowe’s Website Renovation Focusing on Personalization?
Is Lowe’s being timely with its website reno, and is it right to be placed as…
Segmentation is great when (a) segments really exist that are not just a figment of someone’s statistical imagination and (b) it produces groups that are truly different in their attitudes, behaviors, shopping patterns, targetability, and so forth. Nothing says that everything can be segmented in a rational and useful way.
Marketers seem to have an addictive desire to segment, a bit ironic when you consider that consumer purchasing behavior is getting less and less predictable and consumers themselves are falling into multiple segments, often in the space of a single day.
Over ten years ago in our book “The Myth of Excellence,” Fred Crawford and I described the emergence of what we termed the “Instavidual”—people who voluntarily moved from traditional segmentation point to traditional segmentation point. Compound this “instavidualism” with the emergence of multiple platforms, channels and devices and segmentation becomes exponentially more difficult.
Finally, add the fact that lifestyles are changing faster than marketing and you see a generational, ethnic and class disconnect further muddling the segmentation waters.
So, what segmentation can’t do is get to the truth of an individual shopper as a person and—by extension—this means it isn’t going to be too much more successful when it aggregates those individual shoppers.
The only way to get to the truth of the individual is to communicate with him or her, so we should be focusing more on devising better feedback loops and less on which box we can force fit our customers into.
As companies struggle to manage data, it’s good business practice to employ features of segmentation to ascertain and glean insights. Segmentation allows for companies to find a common thread and differentiate information. It’s logical and practical and you can start to understand and pull real truths from it.
But Tyler Anderson is right: it cannot do your work for you. If you don’t understand your market, your segmentation insights won’t mean much. Without a clear brand strategy, what are you segmenting, and why? Segmentation can find commonalities to spearhead innovation, but not create advances. Segmentation is an instrument, along with many other mechanisms, to pool data points and make sense of an ever changing marketplace.
Segmentation won’t take the place of the points outlined in the article, but it definitely helps to make sense of common business questions and finding answers to those problems.
Segmentation is sometimes thought of as nirvana. If I do the segmentation I’ll have the answer. But you won’t have an answer, if you don’t know the question you are asking.
Segmentation research should be designed to uncover growth opportunities. It should be forward thinking not a retrospective on the way life used to be or might be currently. Effective segmentation research requires an achievable objective, a well thought out design and very importantly creative analytics dedicated to the overarching mission of the company.
If the goal is simply to understand existing customers better, the segmentation should allow the retailer to determine what if anything differentiates their high revenue customers from those that spend less money in their stores. That creates the opportunity to develop strategies to generate more revenue from big spenders and small spenders alike.
However, the retailer may have a grander goal. For instance the marketer may want to understand the competitive landscape and segment by consumers who shop in one retail brand or another (or more frequently). That creates the opportunity to generate loyalty among current shoppers and develop strategies to attract new customers.
Bottom line, without a purpose that can drive growth, the segmentation is a waste of time and money.
Segmentation allows retailers and brands to put consumers into convenient buckets for advertising/marketing purposes. The better the definition of each segment, the more tactics can be refined to reach it.
Segmentation needs to go beyond age, sex and geography. It needs to dive into relevancy, usage patterns and frequency of purchase.
Segments need to constantly be reevaluated. Waiting 3-5 years is too long…a lifetime in the digital age.
The ultimate goal for marketers should be to deliver a customized message to each consumer.
You can segment yourself to death these days. I know from work within and outside the supermarket industry that more retailers are dealing with micro-segmentation—groups within the groups.
Retailers should be paying closer attention to the data available to them these days and segmentation is not something you do once for a new store then forget. It needs to be constant and consistent, and I would suggest that every 3-5 years is not enough.
But let’s face facts. Segmentation doesn’t keep the shelves stocked, the store clean, the employees friendly and helpful, or shorten the lines at the checkout.
Marketing quests today still aren’t moving as quickly as the consumers are changing. That can cause an informational imbalance. In dealing with customer segmentation, may we remember that no one has the same fingerprints. True, even as more revealing data is pooled, the ultimate decision by a consumer is still the consumer’s … and thus customer segmentation needs to be used judiciously.
In today’s retail environment, it is clearly not enough to segment. I like the terms used by service designers, ‘roles’ and ‘players’. Every brand needs to understand what the roles and who the players are that come in contact with their brands.
This calls for more than demographic information. It also calls for an understanding of how various consumers interact with the brand. What is the profile in each role? Keep in mind that a single individual may take on different roles dependent on their need at that particular time.
Once the retailer can map this out, the information is golden in how each channel is constructed, merchandised and staffed. IMHO
Just wanted to elaborate a bit re: WayfinD’s contribution here, which I found good, but missing in some key areas:
1. You do not necessarily need to offer niche products and services to avoid monolithic segmentation modeling results. Remember, the segmentation effort is not about the brand or the retailer, not about brands that are offered, it’s about your customers. Monoliths like Hellmann’s, Heinz, most retailers such as Macy’s, Nordstrom’s, Neiman Marcus, can and will engender meaningful, highly differentiable segments of their customer base only if their input bases for segmentation consider and contain many different kinds of input variables: attitudinal, emotional need states, psychographics, lifestyles, shopper styles and behavioral (e.g., situational, ‘moments’) variables. Remember: garbage in, garbage out. You must do the due diligence at the a priori stages (qualitative, planning with stakeholders) prior to the quantitative data gathering phase, otherwise, you will end up with a meaningless effort if your input going in is non-differentiable.
2. Individual resulting segment strategies are perfectly acceptable, given that the common denominator across target segments is the brand’s overall equity, and that this equity be kept intact throughout all strategic development for the individual segments to be targeted. Usually, the strategy is to maintain and enhance each key segment’s equity perceptions, while optimizing each one’s (emotional) brand loyalty, customer satisfaction and retention–all the while maintaining the essence of equity, and all the while attempting to expand core customer usage. The overall goal with each important target segment is to minimize alienation and defection, expand usage and grow the number of customers overall. And yes, in some cases, different strategies aimed at each valuable target segment can be required, as long as each segment strategy is not incongruous with or off overall brand strategy.
3. Mention about the important contribution made by segmentation research to resource allocation is infrequent. Segments that account for high-density, high-index dollar volume spending and consumption in relation to their actual incidence in the brand’s space are considered more important than others. Those that are considered less or non-important will be deemed to warrant less in the way of advertising, marketing and innovation efforts. Segmentation is an important tool to help guide your allocation of dollars.
4. Let the segmentation results and ensuing storylines guide subsequent research (new product concept screening and evaluation, trackers, IHUT product testing, ad communication studies, new packaging and logo design, spokesperson selection, etc.). Make sure that key segment players’ attitudes and responses are represented in such studies by use of the segment model algorithm for study screening and recruitment purposes.
5. Make sure to get corporate stakeholders to buy into the study at early phases, by having them participate and contribute to inputting into the info-seek process. Have them provide their ‘wish-list’ objectives, insights, hypotheses. Where, at the end of the study, do they want to be with this? Having them provide this end-study wish list enables you to track back to insure that this end-goal is reached.
To me, customer segmentation has lost half its value. We used to use segmentation for research and insight, and also to create unique experiences/solutions for each segment.
Segmentation is still a very useful tool for research, and for getting insight out of big data pools, but the days of taking action on specific customer segments is at an end. Technology now lets us take action on individuals rather than aggregated segments of shoppers.
I’m generally appreciative of any discussion that questions conventional wisdom. This thread is a good example.
Segmentation seeks to identify enduring differences between target-able groups. In the early days this was based principally on geo-demographics, later on geo-demo plus psychographics, more recently on behavioral metrics. Some targeting is generally better than no targeting, so these activities have harvested low-hanging fruit for marketers.
Any segmentation, however sophisticated, is at best a snapshot in time. One of the stupendous lessons we must take from SoMoLoMe and Big Data is that consumer behavior reality speeds up and changes much faster than our decision cycles.
So any brand mission statement that begins, “Our customer is…” should be considered suspect if it’s more than a minute old. Three to five years is delusional.
Maybe segmentation needs to give way to snapshot analysis and data mining. That is, periodically look at all available descriptive data about who is buying our product now. Then search the universe of shoppers for others who match those traits and target them.
Ultimately only three segments matter then: (1) The folks who bought your brand this week. (2) Other people who resemble those folks, but who did not buy your brand. (3) Everybody else.
Tyler Anderson raises some very interesting functions and limitations of customer segmentation. Ultimately the strength of any segment lies in how tightly homogenous the segment actually is so that common treatment of each segment member results in relevance and enhanced shopper engagement.
Retailers that still use RFM (Recency, Frequency, Monetary Value) segmentation likely already know that there is much variation in terms of lifestyles, priorities, and product preferences within RFM segments as these segments are devised with just one or two variables. I would encourage any retailer who is still robustly relying on a RFM approach to look at other key determinants of shopping behavior, optimally derived a blend of customer transactional data mining and qualitative research designed to weigh the impact of each determinant on actual shopping behavior.
Conversely, the worst of all segmentation practices results in having shopper segments that are so esoteric and specific that they do not practically lend themselves to action that will benefit the business. I have seen both “polar” extremes in play and neither get the job done.
The online world poses a new challenge: personalization, which in principle works totally reciprocally to segmentation. In segmentation you find similarities among groups of people where as in personalization you are focusing on each person as an individual. However, while creating offers or recommendations, especially in the online world, the struggle would be to draw that line to make it profitable and meaningful for both the retailer and the shopper.
A good segmentation is the one that understands the differences among segments and the reason behind it for the specific event/offer/any merchandising decision. What does not work is the cookie cutter or stereotypical approach to segmentation. Segmentation should appreciate the situational differences; not every individual would respond to a trigger in the same way.
Segmentation can be a daunting task. It can also dilute your brand. Smart brands create a consistent message, even if they are marketing through different channels. They may change the visual aspect, and even some words of an advert, but will still maintain their core message.
Segmentation work is needed in certain industries. For example, a grocery store chain may stock certain items based on the local demographic. A hardware store probably doesn’t stock snow shovels in Miami, Florida, but snow shovels are a staple in Montana. But what probably doesn’t change with the grocery store or the hardware store is their core culture.
For example, if Ace Hardware stores want to be known as the Helpful Hardware place, they are going to helpful in every store, regardless of the demographics.
Put simply, customer segmentation is key to effective promotional planning and price optimization. When you try to create effective pricing for customers as a whole, you’re mixing people who are price sensitive with those who aren’t—and their reactions could be dramatically different to a given promotion. Yet when you measure price elasticity at the customer segment level, not just the product level, your forecast accuracy is much greater and as noted above, you can make strategic choices that provide the most value within your customers segments. This ultimately builds long-term shopper loyalty.
Some great comments below have been made on traditional views of segmentation. I believe the perspective we have talked about for years, that is “A segment of one” is now within our reach. Targeting, marketing, advertising and selling to individual shoppers, as they evolve, and being nimble enough to evolve with them… AND anticipate their evolution, is the new reality with tools available today.
Segmentation is a great marketing tool for both traditional and digital. It is not a silver bullet and does not replace common sense. Segmentation is not everything; it is not a brand strategy nor does it replace innovation or other research. To an experienced marketer, these things seem to be self-evident, however, I suppose they bear saying. Segmentation does provide a solid tool for marketers to be able to understand and communicate with their customers. In the digital age customization is cheap, easy and most importantly expected by customers, segmentation is not only recommended, but almost a requirement.
The idea that a single segmentation base can be usable input for different decisions, i.e. if you are concerned with making pricing decisions, segment on price/sensitivity bases; if a media decision, use listening/viewing/reading, etc. bases. If they’re competitive positioning decisions, another set. Etc.
Customer segmentation is an excellent tool for understanding customers and eventually targeting them. Key is to know for what purpose are you doing customer segmentation. Is it for introducing new products or is it for doing specific campaigns? The way that you do segmentation will change. A unique customer may belong to multiple segments as per the need. Also, as rightly pointed out, segmentation is not a one-time exercise—it changes all the time. What was true 3 years ago may not be close to truth today.
So here’s my 2 cents : (1) Make segmentation relevant (2) Do it every year (3) Couple it with other analysis (e.g. market structure analysis).
The most common mistake I see with segmentation is that it is not actionable. In search of more clever models, we hunt for significant factors and lose sight of the customer.
In marketing, we find the variables that best discriminate are the ones that are most difficult to target, such as attitudes. But the factors easiest to target, like demographics and firmographics, are applied to media buys because that’s where large audiences can be found.
In selling, we borrow from marketing surveys rather than purchase patterns. But the best targets are not those who rate your value prop highly but those who exhibit desired purchase behavior such as R-F-M.
In product development, we aggregate singular preferences into abstract clusters and lose the clarity of individuals’ problems. But the best solution to stale product development is to spend more time with individual buyers
The antidote is to refrain from narrow appeals that come from thin-slicing the data.