• Mark Hallander

Is segmentation a flawed marketing principle?

A core concept in mainstream marketing literature is the process of segmentation, targeting, and positioning. One of its founding fathers, Philip Kotler, argues that marketers must select the most valuable market segments based on demographic and psychographic factors for which to tailor its messages. The STP process is, without a doubt, a very common marketing principle, but have marketers in any way become biased in using it? Is there empirical evidence to support that segmentation is the best practice in any company, industry, or culture?

#segmentation #marketingtheory #heuristics

Challenging segmentation as a marketing principle

In his research paper, Yoram Wind (1978), suggests that the variables used to identify segments are heavily influenced by marketers’ opinions, which makes the process arbitrary. Research has since extended this train of thought by suggesting that segmentation rarely, if ever, has an objective outcome, as researchers make uninformed decisions on different variables.

In relation to this, another segmentation concern is related to the amount of data marketers actually use in the process. While it is fairly easy to segment the market, making assumptions, the lack of concrete data results in poor definition and implementation of segmentation.

Besides this, we tend to assume that consumer behavior is different from one consumer to the other, while it might be more similar than we think. The underlying assumption here is that customers only fit into one segment, while they could potentially be part of a larger and more broad target group.

In the literature, different academics have also raised concerns about the validity of the segmentation process. E.g. the implementation of segmentation strategies (Dibb & Simkin, 1994), whether segmentation can practically work in a market (Danneels, 1996), if segments do in fact exist (Winchester & Lees, 2013) and whether the psychological demographics used to segment the market is stable over time (Hoek, Gendall, & Esslemont, 1993).

"If there is one overriding criticism that can be leveled at market segmentation research in general, it is the failure of researchers to recognize the role of guiding theory in predetermining the general directions and findings of empirical research." – Hoek, Gendall and Esslemont

So where does this leave us?

While these academics suggest that this puts an end to segmenting altogether, I think the implication is instead - much like Marketing Professor, Byron Sharp - that companies are competing in largely unsegmented mass markets. And these markets demand mass communication, more modernly referred to as omnichannel marketing. We are no longer safe, when addressing the consumer via one medium, but should instead think of the different behavior consumers apply in different mediums: Watching TV (commercials), searching online (Google ads), scrolling through social media (Social ads) or finding the right video on YouTube (pre-rolls).

It seems only logical that our brands should be targeted towards specific groups of people, but studies show that user profiles are actually very similar (Kennedy & Ehrenberg, 2001). This goes for both short-term repertoire markets such as radio stations (Winchester & Lee, 2013), and more long-term subscription markets such as banks (Sharp, 2010). And I, myself, have come to learn that market segments are actually much more homogenous than first assumed.

But what exactly does this imply for those marketing practitioners, who seek to create specific target segments in which to position their products?

In my experience, marketing departments still struggle, as marketing managers usually prefer qualitative, experience-driven heuristics rather than algorithmic, data-driven methods. Based on rules of thumb, consumers are fitted into a fixed “box” of predicted behavior - lacking the data to support that behavior. The consequences are severe, as marketing departments struggle to maintain internal legitimacy, budgets are cut, and the influence on c-level erodes (Park et al, 2012).

And what do you do, when you lack marketing effectiveness?

You find ways to collaborate. Media-, dialogue- or digital agencies, for instance, base their segmentation on solid consumer data. Instead of defining subjective consumer segments, agencies will specialize in search or media data, which gives you a clear, quantitative indication of what customers are doing. Through analysis, online behavior for instance becomes a platform for personalized (re)targeting. In that way, these digital marketing agencies also take into account that market segments are not stable, as they change over time.

Is this a sales pitch then?

I don't think so. It is, however, a friendly hint to look carefully at a process, which can become too influenced by gut-feeling and too little based on solid, quantitative data.

This post has only scratched the surface of a real deconstruction of segmentation, so if you want to dig deeper into the research, here are my academic references:

  • Hoek, J., Gendall, P. & Esslemont, D. (1996). Market segmentation: A search for the Holy Grail?, Journal of Marketing Practice: Applied Marketing Science, Vol. 2 (Issue 1), pp. 25-34.

  • Kennedy, R. & A. Ehrenberg (2001). There is No Brand Segmentation. Marketing Insights, Marketing Research, Vol. 13 (issue 1), pp. 4-7.

  • Park, H., Auh, S., Maher, A. & Singhapakdi, A. (2012). Marketing’s Accountability and Internal Legitimacy: Implications for Firm Performance, Journal of Business Research, Vol. 65 (issue 11), pp. 1576-1582.

  • Sharp, B. (2010). How brands grow. Oxford University Press.

  • Winchester, M. & Lees, G. (2013). Do radio stations in New Zealand target successfully?, Australasian Marketing Journal, Vol. 21, pp. 52-58.

  • Wind, Y. (1978). Issues and advances in segmentation research, Journal of Marketing Research, Vol.15, pp. 317-37.