Segmentation: WHO Needs It?

Segmentation: WHO Needs It?

Segmentation can cause contentious debates between Market Researchers.  On one hand, Marketers and Market Researchers who love it, swear by it.  They believe that by deeply understanding which consumer segments they need to target, they can better design products and propositions.  On the other hand, quant-focused researchers like volume forecasters and panel experts are skeptical.  Pointing to the Ehrenberg-Bass Institute’s principles of “How Brands Grow,” these experts, and even high-profile CPG companies, are ditching segmentation.  So who is right?  Or should I say, is WHO right?  If so, how can it coexist with so much contrary evidence?

What is Segmentation:

If you are unfamiliar with the practice of segmentation, it is essentially dividing the market by some set of behavioral, demographic, or attitudinal measures and targeting those consumers.  Often, a market or product category is broken into several individual segments.  Each segment has its own unique traits and desires.  This is used to differentiate products within a portfolio and copy messaging.  If done correctly, it allows a small brand like Sensodyne toothpaste to ring-fence a loyal consumer group of people with sensitive teeth away from mass-market power-houses like Colgate.  Segmentation is a time-honored field of Market Research and marketing strategy.  However, it has many detractors.

For the Love of Segmentation:

It is hard to argue with the expertise of the Ehrenberg-Bass Institute.  Further, it is hard to deny the importance of trial to overall initiative success.  The segment of Marketers and Market Researchers who love segmentation (see what I did there), love it for a different reason.  Many of those who support segmenting their consumers and portfolio due so for the incremental sales opportunities it creates.  After all, segmentation-supporters would say, if you target everybody with every product, you are eventually going to cannibalize yourself.

So Who Benefits From Segmentation:

There are several functions who benefit from segmentation.  Some of these functions are obvious, some less so.  Let’s explore who benefits from segmenting the market and why:

Marketers & Agencies:  Today, it’s hard to talk to everybody with one message in one place.  In the 1960’s when there were only a few television channels, mass-marketing was simpler.  Now, with ultra-fragmented-media and the sophistication of tools like Google Adsense and 84.51 marketers can micro-target down to the individual.  Even if they can’t reach consumers 1:1, they benefit from the guidance of knowing who generally is buying their products.  This enables them to design copy for the audience most likely to buy the product anyway.  By having a segmented portfolio of differentiated audiences, there should be (in theory) less cannibalization across the portfolio.

Market Researchers:  Our job as market researchers is to deeply understand our consumers.  Depth of understanding becomes easier the less broad the group is.  It is easier to have a deep, representative understanding of Millennial Moms than it is all Moms, or all Women in the US, or all Women worldwide.  By working in an organization that values segmentation, Market Researchers play a vital strategic role.  The more granular the segments, the more achievable that role.

Research & Development:  Here again, segmentation gives guidance that makes life easier.  An R&D organization that deeply understands its consumer target can better delight their consumers with great usage experiences.  If they know their WHO and understand her desired product experience, they have a very clear target for what to deliver.  Who doesn’t want that?

Sales & Category Managers:  You might not think that the sales team benefits from segmentation, but they do.  Retailers are interested in growing their category, not individual brands.  If a sales manager can prove that a product reaches an incremental shopper group the retailer isn’t converting, the sale is much easier.

Similarly, Category Managers benefit greatly from having a deep understanding of segmentation data.  Having a clear understanding of shopper segmentation can help them with store clustering.  Further, along with their Shopper Insights counterparts, Category Managers are expected to be objective experts on any question the merchant has.  If the merchant wants to know about a particular shopper segment, the Category and Insights Managers had better have a good answer.

Small Brands:  Like the Sensodyne example above, having a very clearly-established niche segment-target is enough to keep brands alive in the market.  One of my favorites at this game is Moco de Gorilla, who has established tremendous loyalty among US Hispanic males in the hair gel category.

In conclusion, there are many functions who benefit from having a clearly defined consumer segmentation.  Whether it is “just” the data and understanding, or whether it is guiding the strategy for a brand portfolio, there are many practical purposes for segmentation.

Segmentation and the Problem of Real-World Evidence:

So that explains why segmentation is popular.  However, the question is whether or not segmentation is right for your brand.  At a minimum, there is plenty of evidence that it isn’t right in every situation.  Here’s why:

Segmentation Voluntarily Reduces Trial:  In theory, a mass-marketed brand could appeal to every household in the country.  A brand with a specific, targeted segment should not.  In my experience, consumer segments are often sufficiently attractive at 16% to 33% of the population.  Out of the gate, this cuts a brand’s trial potential by two-thirds.

Shoppers Don’t Always Know to Follow Segmentation:  It’s true.  One of the most frustrating things for my former marketing colleagues was looking at segmented panel data.  When we overlaid our segmentation with Nielsen panel, the results were discouraging.  Each of our products was purchased by members of all of our segments.  Granted, some over-indexed with their intended segment, but no product was only purchased by a single segment.  Whether the segmentation was perfectly right, or not, is debatable.  Nonetheless, if segmentation must be perfect to work, that is a limitation of the practice of segmenting as well.  I should also note that not one of those marketers complained about the “extra” shoppers buying their product.

Shoppers Churn:  This is a major problem of segmentation supporters.  What a segmentation strategy lacks in trial, it must make up with loyalty. In theory, this should work since the products and messaging are so precisely targeted.  Unfortunately, research shows that even top brands churn up to 50% of their shoppers annually.  This has two big implications:  (1)  Brands must constantly re-earn trial. (2)  Shoppers aren’t that loyal.  If this is true, voluntarily limiting trial is a bad idea, and loyalty is not a sustainable strategy.

The Problem of Double-Jeopardy:  Lastly, the existence of “The Law of Double Jeopardy” gives empirical evidence against segmentation.  This is another of the Ehrenberg-Bass Institute’s contributions to brand understanding.  Described succinctly:

The Law of Double Jeopardy:  Brands with smaller market shares have fewer buyers, who are also less loyal (loyalty declines with market share).

This is serious problem for supporters of segmentation.  As Dr. Byron Sharp, author of How Brands Grow, observes,

The statistical explanation of Double Jeopardy is that it is a selection effect. Because brand share depends largely on mental and physical availability, rather than differentiated appeals of different brands.  For marketers this is pretty important, pretty insightful, we wouldn’t get Double Jeopardy if brands were highly differentiated appealing to different segments of the market.  Since we do see Double Jeopardy all over the place that suggests that real-world differentiation is pretty mild.  Mental and physical availability must be a much bigger story than differentiation.  That’s a very important insight.

This is consistent with the “Three A’s of Trial” that we discussed in a previous post.

In conclusion, the evidence against the practical success of segmentation is quite compelling.  Does this mean that all the time, effort, and money spent on segmentation is wasted?

Fighting the Urge to Go “Ditch-to-Ditch” on Segmentation:

To answer my question above, no, segmentation is not a per se waste of time.  There is a place for segmentation.  I have worked across a spectrum of organizations whose approaches to segmentation are very different. I have worked in organizations that were designed around their segments.  And I have worked for organizations that wouldn’t even tolerate discussing segmentation.  If I were running a CPG company out of my own garage, I wouldn’t rely on either of those extremes.  [Ed. — And not just because I’d be running a very small CPG company if it were housed in my garage.]

I believe success lies in a balanced approach to segmentation.  I believe this for several reasons:

  • First, I can’t refute (and actually agree) the logic and empirical evidence presented in the Fourt-Woodlock model or in the “Law of Double Jeopardy.”
  • But, there is a practical, organizational need for marketers & their agency partners to select a more narrow consumer target to talk to than “everyone.”
  • In my own experience, I have seen that whether an organization believes in segmentation or not, retailers will want to see that expertise from Category Management and Shopper Insights.
  • I struggle to envision a category portfolio of multiple products that offer different benefits yet simultaneously appeal to everyone.  I also struggle to envision a portfolio that is simply the same product over and over again.  At some point, benefit, experience, and brand equity have to count for something.  Just look at the number of brands in bottled water.  It can’t just be anti-Trust laws that cause all those brands in business.

Therefore, I feel that the best approach is a hybrid approach to leveraging the strategic and tactical value of both segmentation and a mass-marketing philosophy.

OK Plato, Show Me This Republic Where Segmentation and Mass-Marketing Co-Exist:

Alright.  I’ll try.  If I ran the zoo, my approach would be this:

Frankly, I wouldn’t care which consumers buy which products, so long as more of them buy my products more often than my competitors’.  Therefore, I wouldn’t wag my organization or prioritize any other measures than total market share and sales growth.

However,  I would still have research and a point-of-view on the segments.  I’d want those insights so I could talk to my retailers, I would want it to know how to give strategic direction to my organization, and I’d want it to know when/where/how to talk to each individual consumer.

I’d make this investment in segmentation research because I’d need to make decisions within my portfolio.  I wouldn’t want the same technologies, claims, and experiences on my value-tier items as on my premium-tier items. They can share some, but not all.  I want this, for no other reason than the practical reality that I’ll have no other way to justify the up-charge to the retailer or to the consumer if all my products are the same.  Otherwise, I don’t need multiple tiers in my portfolio.  I just need one:  The one with the most benefits at the best cost.

Further, if I have multiple brands in my portfolio, I would want to know what each of them stands-for.  At least I’d want to know what consumers think they stand-for.  As far as I’m concerned, they can overlap and have points of parity.  But if the consumer can’t tell me that each stands for something different, then I don’t need multiple brands.  If they source from each other, fine.  If there is churn between them, fine.  All I care about is that together they reach the largest possible number of consumers.  There must be some reason that Duracell and Rayovac both still exist.  If I owned both, I wouldn’t sunset one just because of Double Jeopardy.  I’d be happy to have that much share of the market.

Lastly, I’d use my segmentation data to figure out how to talk to all possible consumers.  Even if my product(s) is/are one-size-fits-all, I’d still want to put the most effective message for each individual consumer where she is going to see it.  That ability is becoming less and less sci-fi every day.  Therefore, I wouldn’t care if that required a mass-market message on TV.  Or if it meant a different message online, and a different one still on my end-cap displays.  If mental availability is more important than differentiated product appeal, I would want to know where, when, and how to insert my message into every consumer’s share of mind.  That, I believe, is worth investing in — even if I am targeting “everyone.”  Knowing each individual tree is still a way to know the forest.  Prioritizing mental availability is still strategically compatible with Double Jeopardy.


The segmentation debate will still continue on after this article.  The fact that Dr. Byron Sharp’s work didn’t end the debate completely is evidence enough of that.  As a Market Researcher it is hard for me to walk away from the need for segmentation.  I will always want to know what my consumers think.  Since they not all part of the same hive-mind, I have to assume there are differences.

In my own experience, I have seen enough diversity in the world to recognize that different consumers think different things.  However, that doesn’t mean that different consumers can’t think different things about the same product. Perhaps to some consumers a sports drink is healthy.  To others, it’s refreshing.  And to a third group it’s all about taste.  If I can create a sports drink that wins on those three measures simultaneously, great.  I want to know that.  I’ll take those winning perceptions and use them to win share of mind and guarantee share of availability so that I win with as many consumers as possible — no matter why they think they choose my product.


Leave a Reply

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