Conclusion: Dual Chaos
Matching these diverse segments to a broad set of products—in a way that works for shoppers, retailers, and manufacturers—is a “dual chaos” problem. There are a multitude of types and varieties of people (chaos 1), as well as a multitude of types and varieties of products (chaos 2). The question is how to match the people with the products. In the bricks-and-mortar retailing world, it’s not possible (yet) to do an exact one-to-one match. The store cannot be reconfigured to personal tastes every time a shopper walks in the door. As much as retailers might like to customize their stores for every single shopper, this is not operationally practical. So, the best thing a retailer can do is create a “variety” of shopping experiences addressing the distinctive needs of groups of shoppers.
Organizing shoppers into groups is what segmentation or clustering is all about. Although we have considered the three broad segments that have emerged across many retailers, each retailer or store will have more specific insights into how people shop in their stores. There are two general problems of most shopper segmentation. The first is that most of these schemes result in far too many groups of clusters for practical in-store use. Retailers can respond to a small number of large groups inside the store far more intelligently and in a more targeted way than they can to a large number of smaller groups. However, in defense of segmentation schemes producing larger numbers of groups, these may be effective outside the store, where various advertising media may be targeted distinctly to more varied groups.
The second problem is that most segmentation schemes are based on a wide variety of psychographic and demographic data, which although collected by surveys and other research, are not obviously related to in-store behaviors. The goal of the store is to organize the chaos of shoppers into groups and to organize the chaos of products into groups, and then to introduce the appropriate groups of people to the appropriate groups of products. So, in reality, we’re interested in grouping the shoppers by their behavior in the store rather than by their attitudes, opinions, or even need states.
Generally, such characteristics as age, sex, and others inherent to the individual shopper are subsumed. Attitude, of course, is less fixed, but has been given a great deal of consideration in many segmentation schemes. This certainly includes such things as need states and other transitory mental conditions. Although individual characteristics and attitude criteria are of great value in planning outside-the-store communication strategies (advertising), they are more difficult for store management to actually respond to effectively.
Behavior is the critical in-store factor. It is widely recognized that it is more reliable to observe what people do than to ask them what they do. In other words, if behavioral data is available, it will generally be more reliable and relevant than the shoppers’ attitudes and memories. After all, in the end, the only thing that matters is whether the shopper buys—a strictly behavioral matter. Alexander “Sandy” Swan of Dr. Pepper/Seven Up, an early supporter of PathTracker™ studies, once told me: “I don’t care whether the person buying my product is a 60-year-old man who drives up in an $80,000 BMW, or a 17-year-old pierced teen who arrives with her friends in a beat-up VW. All that matters is that they buy.”