Label-Making and the Myth of Market Segmentation

A couple of years ago, in an effort to curb use of disposables around the office, venerable Communispace office manager Janet Toole decided to give everyone in the company a mug and water bottle bearing our logo. Label-makers were provided around the office so we could identify our wares. During a meeting a few months later, I happened to notice that the label on fellow blogger Pete Chapin’s mug was strikingly similar to the one that I—in an admittedly unauthorized use of company supplies (sorry, Janet!)—had recently affixed to my monitor (see picture, below).

What does this tell us (other than that both Pete and I are incredibly witty)? We inherently reject the idea that it’s useful, or even possible, to classify people in simple terms.

I’m a 28-year-old female professional living in Boston. I’ve spent the first six years of my career working at a social media company. From this demographic info, you might assume certain things about me, my lifestyle, my media consumption, my shopping habits, the brands I like and the products and services that might appeal to me. You might even put me into a neat little market segment called “Maturing Millenials” or “Transitioning Twenties.” And, in my case, you’d probably be wrong. In my personal life, I’m not really involved in social networking and prefer my media the old-fashioned way – TV and radio. Yes, real, live radio. Sorry, but I’m missing your banner ad over on Facebook.

That’s just one personal example, but we’re all improbable in our own way, aren’t we? Here at Communispace, it’s not just personal, but business, where labels are concerned. We spend a lot of time—and, at their request, a lot of our clients’ money—finding the right community members based on elaborate, finely-honed market segmentation schemes. Lately, we’ve both been wondering what it’s all worth—and if we’re asking the right questions. In this new world of the Long Tail, niche markets and hyper-personalization, what is the place of traditional segmentation schemes? Is it helpful or even advisable to attempt to segment customers into broad buckets?

We’re not the only ones following this stream of thought lately. Last week I attended an ESOMAR webinar featuring Jochum Stienstra, who noted that, “We can’t sort people like marbles.” In the presentation (based on his paper, “The myth of segmentation, or how to move beyond”) he stressed that, as the predictive value of market segments decreases, we need to let go of strict buckets and take a more dynamic approach to understanding consumers.

We are all unique individuals, and increasingly come to expect the ability to customize and personalize all facets of our lives. So, when marketers fail to take into account the nuances of our personalities, the result is messaging that feels hollow. We see one example of this when brands try to target the ever-expanding Hispanic market. As Communispace’s upcoming research report, “¿Me Entiendes?: Revisiting Acculturation” (due out next month) will show, attempting to define this group based on traditional notions of language and identity fails to take into account the nuances of this very diverse group – but more on that in an upcoming post. And later this year, we’ll be engaging some of our clients in further exploration of their current market segmentation frameworks and how they play out in our communities.

For now, I’m curious to hear from some marketers out there … How does your organization employ segmentation schemes? What successes or challenges have you faced? How do you see market segmentation playing out in the future?

Facebook
Twitter
LinkedIn

7 thoughts on “Label-Making and the Myth of Market Segmentation

  1. Segmentation studies are extremely elaborate and often involved mixed-method quant and qual. Demographics often have little to do with better segmentations, but sometimes play a role in understanding the “now” of a segment. Segmentations have become even more important in a world where category market share is hugely segmented and white space needs to be targeted by new products/marketing. The example of using yourself as someone that cannot be segmented is an incorrect look at how segmentation is approached. You have similarities to other people and segmentation is all about finding ways to target similar people for specific future purposes. Just like we have learned to predict product success with over 95% accuracy, it’s all about finding the insight that doesn’t yet exist. I would love to challenge any market research “guru” to a debate about the power of segmentation studies.

    1. Hi Sean,

      Thanks for your comments and for reading. You’re right – segmentation studies are typically large and elaborate undertakings, which is what’s prompted us and others to want to dig in to what when, why and how they bear fruit. My intent in this post was not to proclaim, or to advocate for, the death of segmentation, but rather to provoke just such debate about the ways that it, like all forms of market research, needs to evolve to stay actionable and predictive. I also understand and agree that certainly not all segmentation is based on simple demographics; I was hoping to represent the perspective of the consumer and not just the researcher in my little example. In any case, it sounds like you are advocating for new and more sophisticated methods of segmentation to successfully target a fragmented market and, there, we can agree. Thanks again for keeping the conversation going.

      (Incidentally, your comment about being able to predict product success with 95% accuracy caught the attention of a few of us here. I’ve never seen that number cited before – can you share a bit about where it comes from?)

  2. Well said Sean! I would add that the ability of a segmentation study to capture consumer heterogeneity is grounded in the design. Also note that the shelf life of a segmentation solution and its predictive value depend on the product category dynamics. Fast changing categories where competitors come in with new products that may redefine the category (e.g. Netflix for video rentals, iPhone for telecom), can make a segmentation absolete pretty quickly. There is also a potential gap between what the research says and how marketers interpret and implement the insights.

    1. Thanks for your comments, Michaela! I think you make a great point that much of the ultimate value of segmentation depends on the design of the study – the methods used and the questions asked – and that is exactly what we hope to explore with some of our clients this year. Not IF segmentation still has value, but how, when and where.

      I also just read your recent blog post on the same topic (http://www.relevantinsights.com/market-segmentation) and think it would serve as a great primer to anyone just getting their feet wet with market segmentation.

  3. Kat:

    Segmentation research is a very broad ranging tool that means different things to different people. At heart, it tries to group people based more on their similarities than their differences, hence your reluctance to be ‘labeled’. It’s not economical for companies to market to individuals, so some level of ‘stereotyping’ is necessary. If labels like millennials are useful to marketers then they will use them.

    I wrote a whitepaper that is featured this week on the BrandingStrategyInsider blog that tries to define segmentation research as a technique to describe what people do and why they do it. There are some practical examples of how segmentation has been applied with good results.

    http://www.brandingstrategyinsider.com/2011/02/defining-segmentation-research.html

    Hope that helps!

  4. Hi Kat,
    I think it really depends on the context – to me that is what determines everything else. There are instances where a simple rule-based segmentation scheme may work as well or better than a complicated segmentation structure based on clustering or machine-learning techniques.

    Also many people associate segmentation with market segmentation. While there is nothing wrong with this thinking, [I believe] segmentation has evolved beyond research and is a wonderful way to garner insights on many fronts (especially in today’s world where there is a deluge of data).

    From a pure research/market segmentation perspective, I agree that experimental design along with certain analytical techniques does play a key role in ensuring the biases etc. are taken care of and the segments are a valid representation of the reality.

    However moving beyond research, meaningful segmentation need not be always complicated to provide insights. A great example is in the area of web analytics. Say your firm wants to increase the number of visitors to your site (acquisition). One can segment the visitors based on traffic sources (search engines, email, social etc.), demographics and some other key dimensions to get some really wonderful insights into who and where they should target for getting value visitors for the best ROI. Segmentations of this nature do not require complicated design structure but is still very useful and beneficial.

    Regards,
    Ned

  5. I work in pharma and undertaking a segmentation study can be a major project. One thing that the team must know in advance of the project is how it will actually be used. From what I understand the answer to this fundamental question is often unknown at the outset of the project. Physician segmentation studies (a hybrid combination of an attitudinal melded with a behavioural segmentation) is an expensive undertaking and, although interesting, if not actionable becomes a worthless endeavour.

    Also, delving into other questions about whether to do a physician segmentation is the fact that medical school and subsequent residency training attempts to homogenize physicians which is obviously an argument against trying to segment them by attitudes and emotions. They are trained not let their innate personality affect the way they treat patients. Therefore, a simple product usage segmentation becomes more applicable.

Leave a Reply

Your email address will not be published. Required fields are marked *