Segmentation
customer segments · segmenting the market
Dividing a market into groups that differ in what they want, what they will pay, or how they buy — so that an offer can be built for one group rather than averaged across all of them.
In practice
Two clients paying the same fee can be entirely different businesses to serve: one wants the work done and left alone, the other wants to be involved weekly. Pricing them identically means one is subsidizing the other.
The common mistake
Segmenting by what is easy to measure rather than by what changes behavior. Company size and industry are convenient and frequently irrelevant; the segment that matters is usually defined by a situation, not by a demographic.
A market is a collection of people with different problems who happen to be buying in the same place. Segmentation is the decision about which differences matter enough to act on.
The reason it is worth doing is arithmetic rather than marketing. An offer built for the average of a market fits nobody in it, prices at the average willingness to pay, and is beaten on both ends — by a specialist for the demanding segment and on price for the undemanding one. Averaging is the structurally weakest position available.
Segmenting by the wrong thing
The usual segments are size, industry and geography, because those are the fields available in a database. They are sometimes right and they are chosen for convenience, which is a bad reason.
The segments that pay are defined by situation: what the buyer is trying to get done, what happens if they do nothing, who inside the organization feels the problem, and what they have already tried. Two identically sized manufacturers, one replacing a system that just failed and one running a planned review, are in different markets. They have different urgency, different budgets, different decision processes and different tolerance for risk, and no demographic field distinguishes them.
The test for a segment is not whether the members resemble each other. It is whether a change in the offer would move one group and not the others. If the same proposal works equally well across a boundary, the boundary is a category, not a segment.
What it is for
Building an offer that fits. A narrower group has more consistent requirements, so the thing can be made more specific, which is where differentiation actually comes from.
Pricing differently. Segments with different willingness to pay can be charged differently without either feeling cheated, provided the versions genuinely differ. This is price discrimination done legitimately.
Deciding who to refuse. The most valuable output and the least used. Segmentation that does not result in a group you stop pursuing has not been acted on.
Knowing which numbers to trust. Unit economics averaged across mismatched segments are misleading in both directions — the good segment looks worse than it is, the bad one looks survivable.
How far to cut
There is a real tradeoff and it is not resolved by cutting as finely as possible. Every additional segment costs something: a separate offer, separate material, separate delivery patterns, and a sales conversation that has to identify which segment the buyer is in before it can proceed.
A practical limit is that a segment should be large enough to be a business on its own and distinct enough to need its own answer. Below the first, you have built a bespoke offer for a handful of clients and called it a strategy. Below the second, you are maintaining two versions of the same thing.
Most small businesses are under-segmented rather than over-segmented, and the reason is not analytical. It is that choosing a segment means declining the others, and declining revenue is difficult in a way that no framework makes easier.
Concept web
Open the full webQuestions
What is market segmentation?
Dividing a market into groups that differ in what they want, what they will pay, or how they buy, so an offer can be built for one group rather than averaged across all of them.
What makes a good segment?
That a change in the offer would move one group and not the others. If the same proposal works equally well on both sides of a boundary, the boundary is a category rather than a segment.
Why is segmenting by company size or industry usually wrong?
Because those fields are chosen for convenience rather than because they predict behavior. The segments that pay are usually defined by situation — urgency, what has already been tried, who feels the problem — which no demographic field captures.