Let Segmentation Go Deeper

Consumers may differ in their demographics, lifestyles and personality, tastes and preferences, attitudes and behaviour. Consequently, a couple of consumers who differ on any of these characteristics may experience differently the same event or activity, service episode or product usage (e.g., by the thoughts and conclusions, feelings and emotions the experience evokes in each consumer). The history of past personal experiences is also likely to shape the response to a new experience of a similar kind. A segmentation scheme has to be multi-faceted, fine-tuned and sensitive enough to capture the possible distinctions in personal experiences and consumer responses to them.

Gross or too generalised segmentation variables may not be concrete or sensitive enough to identify relevant differences; using a single segmenting variable or bi-dimensional segmentation would often also not suffice to distinguish between personal experiences of consumers and anticipate their likely responses (e.g., to a service call or advertising message). Demographic characteristics particularly tend to be less sensitive and indicative than the other types of characteristics mentioned above. But even when using those psychological and behavioural characteristics, we would need to include more types of them, allowing for greater variety of consumer data, to tap how consumers are likely to respond to specific personal experiences.

Priyanka Carr, COO and formerly general manager of market research at experience management firm Momentive, is advocating the utilisation of detailed segmentation as opposed to reliance on aggregate data and shallow segmentation (“Understanding Individual Consumer Experiences Matters“, Quirk’s Media, 14 June 2021 [registration required]). In her article for Quirk’s, Carr explained, through examples, the mistakes that can be made by not accounting sufficiently and properly for variability (heterogeneity) in consumer characteristics. It can lead, for example, to missing out on more intricate differences in consumer preferences, and thus might not recognise the difference in appeal or meaning of product versions to distinct segments. By not going into sufficient and relevant detail in choosing the bases of segmentation, one may also fail to identify sensibilities such as biases and stereotypes, which could turn into disturbing pain-points for consumers and brands. Using two or more demographic variables can already make an improvement over reliance on a single demographic criterion, and yet it is advisable to incorporate more types of variables or bases for segmentation other than demographic to detect crucial consumer differences.

Carr highlights the need for “solving for what’s not obvious” as the most impactful way for generating meaningful and actionable next-level research. She proposes tips for six steps that marketers and market researchers can take when conducting segmentation research: (1) Embrace buyer personas; (2) Avoid analysis pitfalls; (3) Add questions about basic demographics; (4) Be thoughtful about demographic questions; (5) Collect segmentation data; and (6) Use high-quality, representative samples.

Henceforth we will focus on Tip 1 regarding buyer personas. Carr restricts her examples to demographics, apparently to keep them clear and simple, but her recommendation to construct and employ buyer personas suggests that she aims higher, or rather deeper. That is because the ‘persona’ is a construct that incorporates psychological and behavioural characteristics for creating a richer, more vivid and approachable image of a consumer exemplar. The first tip on personas is explained in full as follows:

Embrace buyer personas, which are detailed descriptions of target customers based on customer and consumer research. Buyer personas let companies talk to customers about their products in a personal way that highlights the benefits that are important to the personas.

Let us develop this idea a little further with respect to consumer segmentation. A researcher may start by taking a look from high above the field, that is by using more general consumer descriptors (e.g., demographics) to reveal the more fundamental patterns of differences among consumers. But these patterns probably would not relate close enough to specific consumer experiences, responses, or behavioural tendencies in actual situations — the more gross model is meant to serve the researcher just as the baseline or template for deeper, more detailed enquiries about consumer or customer differentiators. The questions from this stage onwards are to what level of depth or detail to go, and how to do this.

Categorization tasks are ubiquitous in human thought and behaviour. A category may be described by rules (i.e., based on object features), or it may be represented by exemplars (i.e., an object image as ‘compilation’ of features). Subsequently, people may approach a task of assigning a new item to the appropriate class by testing what rule(s) applies to the candidate item or by comparing the candidate item to exemplar(s) to see to which it resembles most closely. An exemplar is often regarded as a prototype of a category or class. The profile constructed for a persona (of e.g., buyer, consumer, customer) constitutes this kind of exemplar.

The task of the market or consumer researcher is to construct a number of personas as exemplars representing different types (segments) of consumers (or buyers). The persona has to be believable, resembling a real consumer, though it does not have to be a real person (and ethically it should not be so). The profile for the persona may be constructed, for instance, by integrating a few more typical consumers in a segment, or it may be based on a particular representative consumer and then augmented with some additional attributes also strongly associated with a segment to create the final persona exemplar. Personas may be characterised by their lifestyles (e.g., how they allocate time between work, leisure and family), shopping habits, personality traits (e.g., open-minded, innovative & explorer), and more (this can be linked with Tip 5 on collecting segmentation data). Marketers can develop and design product variants or service formats to fit any of the different personas identified or create different versions of media content that would appeal to them better. Thereafter, the marketer has to target real-world consumers according to their closest association with the exemplar personas (i.e., where they might be found and how to approach them).

It should be noted that advanced personalisation technologies allow companies to identify and profile actual individual customers based on various demographic, behavioural and psychological characteristics, then making them personally-fitting offers, and even creating personally-customised versions of product models for individual customers. However, this approach is not feasible for every product or service and market situation. In many cases going through targeting based on personas is the more practical and economic approach. Nevertheless, even when personalisation and customisation are technologically accessible and feasible, the personas may serve as a helpful intermediary, like a guiding tool that focuses and refines the targeting effort.

Deeper segmentation can give greater power to companies and their marketers to approach consumers with better-fitting and more appealing offerings, service & support, and media content. The bases of segmentation should aim to be more detailed, fine-tuned to the anticipated bahaviours of consumers-customers (i.e., achieving higher relevance), and be of greater variety (i.e., incorporating demographic, psychological and behavioural data). Consumer (or buyer) personas, applied as segment exemplars, can serve as the tool for making segmentation models practical, and for guiding and refining targeting efforts.