If you think about it, the very notion of segmentation is quite weird. Market segments are supposed to be specific sub-groups of customers sharing some common traits or qualities that will behave the same in regard to your product.
It's convenient for marketers as it makes their work as simple as possible: customers in Segment A get to deal with product A at price X, while customers in Segment B get to deal with product B at price Y... From there, you can unroll targeted promotional campaigns for each segment. Even better? Insulate each activity in its own product line!
On top of my head, here are a few reasons why this (mostly) never works:
- Boxing customers into segments reduces them to a few attributes that might be relatively stable over time or rapidly changing. However, cultural trends and shifts always transcend traditional market boundaries.
- Each attribute that you pick to decide if a given customer is instead in segment A or B usually oversimplifies reality and, worse, often masks where the real market drivers are (don't get me started on packing every generation as GenX, Y, or Z to describe how they behave).
- Even if you get it somehow right, segmentation always lags in detecting emerging market niches or trends outside predefined attributes. Grand theories and processes exist about segmenting markets that "don't exist yet," but scratch the surface, and you'll soon realize you're practicing astrology rather than business.
- Even when the market exists and key attributes defining a segmentation seem to be working, the speed at which these attributes will change is always grossly underestimated. They soon evolve into deeply entrenched cultural bias and static corporate fiefdoms in charge of aging product lines.
- This leads to a classic innovator's dilemma as, by nature, market changes very rarely respect the boundaries of your current business units. Emerging needs and problems to solve will mostly be around a cross-section of your previous segments, with different categories of customers overlapping together in ways you won't want to deal with, as this will mess up your power hierarchy.
- When innovation is driven by market changes (socio-demographics, behavioral, economics, politics, etc), there's still a chance to solve some of the aforementioned issues, as you might be a wizard capable of finely monitoring how your market's attributes are evolving. But when innovation comes from the rapid introduction of new technology, this goes out the window. There is very little correlation anymore between the set of attributes you're using and how/when the market will adopt or not this tech. (The notion of 'digital natives' is a famously terrible example of forcing attributes on a market that doesn't behave as you'd want it to.)
- From an external perspective now, overemphasis on segmentation will also result in misreading how rival companies move in and develop, as their strategy will rarely conform to your traditional market segmentation. (This is quite the understatement as 80% of business literature is about outthinking competitors with 'blue ocean' or 'white space' strategies, i.e., rejecting consensual segmentation.)
But the top cardinal issue with market segmentation often comes from a core issue: thinking in terms of products. Customers don't care about your products; they care about the problem they want to solve with (maybe) your products.
You can, for a while, maintain the illusion your segmentation matches the market's behavior, but you should know that, at best, this is an odd event in time. And that's the real dealbreaker for most segmentation strategies: trusting there is strong causation between a product line and a set of customer attributes is rarely verifiable, if ever. There might be some surface-level correlation from time to time; it might even be significant for a period of time, but it will never last as you're missing the real market drivers.
What are the problems to be solved? And how should your organization reflect this? The answer is rarely siloing product lines for the next five to ten years.
