There are lots of different ways to segment markets. Many of them are useless.
Market segmentation draws on a simple, yet compelling, core premise: Consumers aren't all the same, so they shouldn't be treated identically. Marketers recognized this fundamental fact a very long time ago. Yet the issue remains as relevant today as it was when Wendell Smith talked about it half a century ago. Surprisingly, however, the past 50 years have produced precious little progress in how companies think about, define, and deal with consumer differences.
Consumers certainly differ in any number of ways. For marketers, some differences are important, but many are trivial. Separating the important from the trivial is key, since only the important differences have a direct impact on the effectiveness of a company's marketing programs.
The solution sounds so simple: First, identify the important determinants of consumer response to a company's products or services, then segment consumers into buckets accordingly. But exactly what are these "important determinants"? What should be the basis for segmentation into "buckets"? That's where the battle begins.
Slicing segments for success
As any methodologist would readily affirm, there are vast numbers of ways in which consumers can be divided, grouped, clustered, or classified into segments. These range from simple crosstabulations and correlations to far more complex factor, cluster, and latent-class analyses. Regardless of the method used, however, the logic of segmentation remains essentially the same, and so does the goal: making marketing more effective by identifying a meaningful and manageable number of distinct "people types."
However, in spite of the ready availability of segmentation methodologies and the obvious appeal of -- and need for -- dividing markets into meaningful groupings, there has been a disconnect between segmentation's compelling promise and its typical performance. Yes, there are cases in which market segmentation has helped focus product development or marketing and communications planning. But there also have been far too many occasions in which segmentation studies have produced hugely disappointing results.
The key consideration isn't how to conduct the clustering -- but rather, what basis should be used for clustering and grouping consumers. Merely defining some number of consumer groupings doesn't ensure that the identified segments will prove to be useful in helping to focus a company's marketing efforts. Tossing myriad measures into the statistical hopper and hoping that the computer will somehow discern a meaningful pattern probably won't yield valuable marketing insights. Nor will it show the way to a more effective marketing program. A million monkeys and a million typewriters have yet to produce another Hamlet.
The real challenge is to identify meaningfully distinct segments of prospects and customers. After all, the goal of segmentation is to make companies smarter and their marketing efforts markedly more productive. With this in mind, consider the multitude of measures that have typically been used for segmenting prospects and customers, including:
- affective measures: attitudes, images, and opinions about brands and brand-related activities. Important and useful? Not always, since many seemingly relevant attitudes and images have proven to have little or no direct relationship to what customers actually do.
Consumers who hold a brand in high esteem may nevertheless choose not to purchase it. Satisfied customers may switch allegiances. Brands with a great deal of cognitive or rational appeal turn out to have no emotional appeal. Attitudes, quite simply, are imperfect predictors of the business outcomes for which brand managers are justifiably held accountable.
- behavioral measures: current category and brand purchase/usage patterns and experiences; communication and shopping behaviors. Useful? At times. Yet current purchase and shopping habits are often inadequate predictors of future behaviors.
Current behaviors may have been bought (or bribed) at great cost -- and with no lasting benefit -- to the company. Also, the company's greatest opportunities may not lie with heavy category buyers or users, in spite of the appeal of "fishing where the fish are."
- classification measures: geographic and/or demographic characteristics. These are readily measurable, of course. But there are far too many examples of older consumers who behave like younger ones, strugglers who behave like affluent consumers, and near neighbors who nonetheless exhibit completely different shopping and purchasing behaviors. Demography cannot separate great prospects from mediocre ones.
- general measures: global life stage and/or lifestyle attitudes, interests, and opinions about life, family, and life requirements. These are tantalizingly interesting, to be sure. Yet the track record for lifestyle segmentation's usefulness is, at best, mixed.
Segmentation on this basis assumes that the defined groupings are somehow enduring. They are not. Situational variables intervene. Consumers are not the same in all situations. A consumer may well be a "coupon clipper" when it comes to buying laundry detergent and a "sybarite" when it comes to selecting perfume or a bottle of wine.
product-category specific measures: product attributes, benefits, wants, and needs. These most directly pertain to the company's particular brand and product mix, and thus hold great promise of actionability. However, in practice, the focus tends to be on the rational product/service attributes or on rational wants and needs that will supposedly differentiate competing brand users/prospects and help to define targets for each. What's wrong with that? Merely the fact that consumers do not select among alternatives -- most of which increasingly offer parity of performance -- based on rational considerations. The rational side of human nature reveals only a portion of the true picture, and it can be greatly misleading.
Companies have relied on these various measures, sometimes separately and sometimes in combination, in their attempts to detect the clusters or segments that might exist within the audiences to whom they market. Each of these measures, however, has fallen well short of meeting the overall challenge to be useful. They have either proven to be unstable, non-actionable, or irrelevant.
Something is clearly missing.
One major problem is that so many segmentation bases are in fact only minimally related to actual business outcomes. They may appear interesting. They may even appear intriguing. They generate lots of workshop discussions. But in many cases, they are also largely irrelevant. They result in segments that either can't be reproduced, can't be addressed, or have no marketing viability. Thus, they are of no help to a company seeking to improve its marketing planning and performance.
The net result is that far too many applications of segmentation, as Clayton M. Christensen and Michael E. Raynor note in The Innovator's Solution, result in marketers designing "me too" promises and products that merely mimic the features and attributes offered by their competitors. And, we might add, they also result in "me too" marketing that targets the exact same audience "buckets" with the exact same appeals. Thus, instead of stimulating differentiation in product performance or brand communications, the result has been quite the opposite.
Segments that matter
What's missing from the extensive list of alternative segmentation bases is this: something that can generate the crucial marketing insights that simple geographic, demographic, or behavioral segmentation schemes lack, something that truly reveals key insights about consumers.
Too many approaches reveal only superficial information. Too many create the misleading impression that companies know their customers simply because they have applied human names or psychographic labels to consumers -- clusters that may be neither enduringly real nor meaningfully distinct. Naming is not the same as understanding.
What has been missing from each of the available approaches is a deep understanding of the ways in which consumers form lasting connections with the brands they buy and use. Scientists at Gallup, however, have been probing the nature of these relationships through extensive exploration into "customer engagement" and the bonds that are forged between companies and consumers. (See "Connecting With Consumers" and "Building Brand Differentiation" in See Also.)
These Gallup investigations have revealed a simple truth that sheds important light on the ways in which companies can identify and act upon truly meaningful customer segments. The connections, and the potential for connections, that exist between consumers and branded products or services are, at their core, emotional. More to the point, these emotional connections represent a powerful new way to segment consumers -- one that offers companies a more complete view of their challenges and opportunities.
- Emotional connections are meaningful. Gallup researchers have shown that these connections are bellwether business performance indicators, directly related to key business outcomes ranging from share of wallet and purchase frequency to customer retention and profitability. Customers who are fully engaged with their current brands represent assets that are far more valuable than those who are less engaged. Prospects high in "engagement potential" for a brand are far more likely to form a lasting relationship with that brand. In short, engagement matters.
- Emotional connections are actionable. Gallup has shown that the key marketing drivers (product, promotion, process, price, place, and people) of these emotional connections can be reliably identified and, more to the point, they can be enhanced by marketing actions.
- Emotional connections help define the extent to which alternative consumer segments are viable, because these connections reveal the extent to which lasting consumer relationships are being built. Heaviness of use does not necessarily indicate the strength of the customer relationship, nor is it a useful tool for targeting desirable prospects. Heavy users may switch brands, and they may do so regularly. Emotionally engaged users do not.
Impressive credentials, to be sure. But why would consumer segmentation that is based on a brand's evident degree of "emotional connectedness" be a better solution for guiding a company's marketing efforts?
Of course, it could be interesting for a brand manager to know that there is a large group of consumers who are 35-49 years of age and earn in excess of $75,000 a year. It might also be interesting to know that they have written a letter to their state senator, shop for clothing at least once a month, associate a particular store with having a wide variety of current fashions, say they try to keep up with the latest fashions, and reportedly choose their clothing based on how well they feel the clothes are made. But is it really useful to know this? In truth, not one of these measures reflects the extent to which these consumers have formed any sort of emotional bond with a store.
How much more useful would it be to know the extent to which these consumers have an emotional connection with a particular store -- but not with its major competitors? Only emotional connectedness represents the existence -- or the lack -- of a meaningful bond between the brand and the consumer. And yet that brand connection represents the company's most important asset.
Segmentation based on measures that reveal the true strength and health of a company's consumer relationships provides companies with the crucial insights that have been missing from their previous segmentation efforts. Whether measures of emotional engagement are focused on current customers, as measured by customer engagement, or on non-user prospects, as measured by engagement potential, they provide the basis for market segmentation that can at last meet the management challenge.
How important is an emotional connection? Nobel laureate Daniel Kahneman contends that emotions have a basic and elemental role in all decision making. Marketers who fail to include the essential ingredient of emotional connectedness in their segmentation efforts will find the resulting consumer portrait incomplete or even misleading. Without a fundamental focus on what it takes to create enduring consumer relationships, the promise and power of segmentation will remain unfulfilled.