27: Market Segmentation
From a biological perspective, all humans have essentially the same brain; but beyond the basic motivations for survival, the way in which people think and act is highly idiosyncratic. Differing cultures, personalities, and experiences result in different responses to stimuli. Two people may do different things in response to the same stimulus, and even when they do the same thing, it may be for different reasons.
In marketing terms, different people will buy different brands for different reasons, or the same brand for different reasons. Their reactions to packaging, advertising, and all the various tactics that are used to lead them to a given brand will likewise differ. Considered in terms of heuristics, each person has their own set of questions and requirements. Even when the use-based needs are similar, the structure of the heuristics will be different, as the data (knowledge/memories) that drives the heuristics will be different.
As such, where there exists more than one brand in a marketplace, each consumer will choose the brand that best satisfies their personal heuristics, and best speaks to their idiosyncratic way of thinking. Except in instances where, for whatever reason, only one brand is available, people will choose different brands, and there never has been an instance in which 100% of a market, without exception, chose one single brand.
All of this is a seemingly insurmountable challenge to marketers. Whereas a salesman, dealing with a single client, can choose a tactic and then adapt his presentation to the idiosyncrasies of an individual customer, a marketer prepares a single presentation to be delivered to all consumers, and is unable make adjustments on the fly.
As such, the best a marker can do (EN: or "could do" in the age of mass media) is to focus on a smaller group of people, whose thought processes and heuristics he assumes to be largely similar, such that a single presentation will be positively received by a significant portion of that group.
Given the response rates to advertising, even advertising that seems to be focused on a very specific and well-defined segment, are dismally low, the segmentation approach to marketing has been a consummate and consistent failure.
It is believed that this is due to a flaw in the marketing process - if we could work out the bugs in a segmentation strategy, it would work flawlessly, or at least better than in presently does. This seems credible because segmentation itself is done so poorly: by arbitrary decisions, gut feel, and little formal market research (though larger companies have the resources to do a better job of it than smaller ones).
Historically, segmentation began with simple demographics: young people or old, the affluent of the mass market, male or female, etc. (EN: I would argue that it started earlier, with the geographic regions served by a firm, or the reach of specific media channels, but given the author is likely to shift gears soon, I won't belabor that point here.) They soon realized that even within a given market segment (wealthy young males), consumers still used a range of different brands - a clear indication that demographic segmentation doesn't quite work.
Stubbornly, marketers refused to accept the flaw in the practice, but rather decided that it was not a problem with segmentation, merely that the variables needed refinement. So they tweaked their equations, not only refining demographic characteristics but also adding a few other factors that were facile: hobbies and interest, living standards, political beliefs, and the like. This resulted in marginal improvement.
Various schemes for improving segmentation followed, which brought on the advent of onerous marketing surveys. Researchers attempted to ask as many questions as possible, including very intrusive and personal questions, because they had no idea what qualities should be used to define a segment. They also attempted to load many product categories into a single survey to get as much data on as many products as possible, which had a multiplying effect on the length of the questionnaires (twenty questions about twenty products yields a 400-question survey).
This also led to the specialization of marketing firms into niche markets: a firm might claim to have particular insight into affluent customers, or customers in a given nation, or customers of a given race, or customers who are members of "Generation Y." None of this addresses the simple truth:" that segmentation doesn't really work all that well.
What's more, segmentation studies do not get to the heart of the matter. The critical question is "why do people choose our brand?" Statistical analysis of the consumer base, showing correlations between given segmentation metrics and the body of consumers, will tell you what "kinds" of people happen to be using your brand today - but offer not a shred of evidence as to the reason they do so.
Even so, the trend of segmentation continues: it provides little useful information, but it results in a marginal improvement in advertising response rates - and so long as there are enough customers in a given segment, it's likely perceived as profitable to continue for firms to utilize it, and for agencies to leverage it.
Segmentation and Secondary Needs
Studies into consumer behavior indicate that actual purchasing is chiefly driven by two factors. First, consumers seek products to satisfy their needs. One may debate the validity of a need, but in the perspective of a given customer making a purchasing decision, need is involved as the primary motive in a purchasing decision. Second, once a need for a product is accepted, customers are driven by other factors (culture, identity, ambition, etc.) to select an option among those that are perceived to satisfy the need.
As such, marketers should seek first to identify a segment of the market that has a need that can be solved by their product, and then to decide whether the brand should be positioned to address the secondary interests of the consumer. Theoretically, it's possible to market to the secondary interest while ignoring the primary interest of need, but the author "cannot think of an example" where this has been done successfully.
He uses the example of cornmeal in South Africa, where finely-ground white cornmeal was considered to be the best quality (white corn being more rare than yellow, and fine being harder to produce) and served to the nobility (king and court) and honored guests. To satisfy the need of hunger, fine white cornmeal and coarse yellow cornmeal are equally effective; but the secondary interest (social status) leads to a preference for fine white. This preference persists even among the urbanized population who have never lived in the kraals.
In the same markets, there are many brands of cornmeal (it being a staple food): yellow or white, coarse or fine, enhanced with vitamins, sponsored by poplar athletes, positioned as traditional or modern, promoted in English or the local dialect, some offer the best price, some claim to be more convenient, etc. None of these qualities that define a brand have anything to do with the primary need it provides (hunger, nutrition).
As such, the primary need served by the product remains the same across brands - and all consumers are striving to satisfy the same basic need. The differentiation, the desire to appeal to a secondary need, is what distinguishes one from another, and makes one brand more appealing to some customers than others.
(EN: One thing to note is that all of this seems to arise from the situation of a marketer who is given a product and expected to define and reach an audience. The alternative approach, to survey the marketplace to discover unmet needs and design products to meet them, is not considered.)
Some Suggestions About Segmentation Studies
If there are multiple brands in the market, and different people have a preference for one of them, then they have by that behavior segmented themselves. Clearly, there are brand X buyers and brand Y buyers - and the question is: why do people choose the brands they do? If a given schema fails, it is because that question has not been accurately or adequately answered.
As such, when people proclaim that segmentation "does not work," there is the distinct possibility that it is because the segmentations schema did not prove to be true, or that there was some execution problem in the study. The fact that people tend to favor one brand over another (for a given purpose and occasion) is indisputable. The reason why they do so remains a mystery.
The inherent problem with the parameters that market segmentation considers is obvious: a person who prefers a brand can be described by ethnicity, gender, and age - but these qualities are entirely incidental. Ask a person why they prefer a given brand, and they are unlikely ever to respond that it is "because I am a 30-year-old middle-income Hispanic mother of two young children."
That's not to suggest that there is no correlation between brand preference and demographics, merely that the connection is tenuous. For example, we can correlate age to preference, but this is likely not a direct determinant, but a common factor - a given brand may have been fashionable at one time and younger people were not exposed to the same experiences (advertising or otherwise) that caused older people to form a preference for that brand. Or we can correlate a preference to a given brand by parents of young children, not simply because they are such but because they give greater consideration to nutritional value than to flavor.
(EN: it occurs to me that the notion of segmentation becomes a self-fulfilling prophesy. If we believe our brand is suited to a specific market segment, we place our marketing and logistical efforts into supplying our brand to that segment, it increases that segment's experience with our brand and, lo and behold, we then observer that segment has greater affinity for the brand than other segments, whom we have given less exposure and have even told them that "our brand is not for you.")
Another problem is that research follows the academic discipline: the first step in any study is to conduct a review of literature to determine what has already been discovered by others and build upon it, rather than re-testing what has already been proven. This is regarded as a sound practice, but results in conformity: not only do we automatically accept as true what others have proven without much consideration of their methods, but we also come to accept the methodology as being valid and design any new experimentation according to the same premises and practices.
With that in mind, the best way to approach a segmentation study is not to consider, as has been done repeatedly in the past. the various descriptors of the people who use a brand, but by considering the various reasons any consumer might choose to use the brand.
Once segments based on choice are determined, we can then consider the various other qualities of the audience, and we will likely find statistical correlation to the various demographic factors to define an audience - but this is less in the nature of creating a profile of the user than it is to implementing strategy: it is not significant that a person of a demographic profile uses our brand - but when we discover the reasons our brand is preferred, and discover that may of the people who consider this reason to be of a given demographic segment, we must then plan a way to reach that segment (the places they shop, the media they consume).
There is the caveat that shopping behavior and media consumption are separate issues about which many questions can be asked - resulting in a terribly long and onerous survey. For a brand marketer, focus on the reasons people choose a brand, and leave considerations of shopping habits and media consumption to separate studies.
Neuromarketing And Segmentation
Neurology focuses on the function of the human brain, an individual organ that motivates an individual person. There is no collective brain, nor any research design that can be utilized to observe the effects of a stimulus on multiple people at once. Segmentation is a method of analysis, of aggregating and classifying the results of many individual studies to observe commonality.
Even at that, we can observe that different subjects feel differently about a given brand, or even feel the same about it, for different reasons. The results of a brain-scanning experiment will show a positive or negative reaction in an individual exposed to a stimulus - but we are in essence noting what parts of the brain "light up" without having any indication of the reason they are doing so.
Simply dividing the subjects into groups of users and non-users is "not a good experimental design." To understand motivation, we must understand the heuristics they utilize, and seek to have specific understanding of where the "dopamine moment" occurs for a given product category, and how brands in that category are differentiated.
The author considers a research project that two of his students are planning (but had not yet conducted at the time this book was written) into consumer preferences and chocolate - a product that has minimal utility (it is not sought to remedy hunger or gain nutrients) and therefore is driven largely by emotions.
The target market for nearly every chocolate or sweet confection is females - which raises certain questions: is there a difference between female and male brains that would explain this? But gender aside, there are very different behaviors evident among market segments:
- People who simply cannot resist sweets, whose behavior is similar to that of a drug addict
- People who see sweets as a way to reward themselves for doing something
- People who see sweets as a high-energy good that has a practical purpose (to improve their alertness or provide fuel for physical activity)
- People who are merely self-indulgent, for whom sweets are a means to experience pleasure for its own sake
- People who purchase sweets as a gift or reward for someone else, and do not consume it themselves
- People who do not purchase sweets, simply because they do not enjoy them
Each of these segments will have different reactions to a the product itself - they will respond or not respond regardless of brand. That is, the person who seeks chocolate as fuel for physical activity will have a different reaction to someone who seeks it for reasons of pleasure.
The point the author is attempting to make by this example is that the motivation of the individual, the heuristics they apply, will have a significant impact on their reaction in terms of an MRI or EEG study, which will detect the positive response, but will not indicate the reason behind the phenomena observed.