Marketing to the Average: The Customer You Designed For Does Not Exist

In Brief: Nobody is average. And a brand built for the average customer is built for someone who does not exist — which is why it struggles to be the most relevant choice for anyone.

The average person has one testicle.

The line is crude but it makes the point faster than any statistic. Average the population and you get a person with one testicle, one ovary, and slightly under two legs. The average is real as arithmetic. It corresponds to nobody.

There is a more serious version of this, and it cost the US Air Force a great deal before anyone noticed.

In the late 1940s, American fighter pilots were crashing at an alarming rate with no mechanical explanation. The cockpits had been designed in the 1920s around the average dimensions of the pilots of the day, and the assumption was that pilots had simply got bigger. So the Air Force commissioned new measurements — thousands of pilots, ten physical dimensions each, to recalculate the average and redesign around it.

A young researcher named Gilbert Daniels asked a different question. Not what the new average was, but how many pilots actually matched it. He took the middle 30% range on each of the ten dimensions and counted how many pilots fell inside it on all ten.

Zero. Out of more than four thousand.

Not a single pilot was average. The cockpit had been designed with precision for a man who did not exist, and every real pilot was fighting it.

Most brands are still building that cockpit.

The average customer in your category wants moderate quality at a moderate price with moderate convenience. Build for them and you produce something that fits nobody particularly well. Which leads to "meh" brands. Unremarkable. Boring.

It is not that the proposition is wrong. It is that it has been optimised for a statistical artifact rather than for a person — and people are who make the choice.

This is how brands end up acceptable to many and compelling to none. The positioning is broad enough to include everyone, which means it speaks to no one in particular. The proposition covers the full range of use cases, which means it excels at none of them. The identity is carefully calibrated not to exclude, which means it attracts nobody strongly enough to defend it.

Strategy, after all, is about making specific choices that lead to specific people choosing specific choices.

Every one of those decisions was made to expand the addressable market. Collectively they produce a brand nobody has a strong reason to choose.

There is a legitimate counter-argument, and it deserves stating properly before dismissing it.

Some brands genuinely should speak broadly. Coca-Cola is not trying to be the most relevant choice for a specific segment — it is trying to be available to everyone, everywhere, all the time. In categories with low involvement, low differentiation, and enormous scale, penetration logic dominates and broad appeal is the correct strategy.

But that strategy has a prerequisite most companies quietly ignore: you have to already be the leader, with the distribution and the media weight to sustain it. Broad appeal is what market leadership looks like from the outside. It is not the route to getting there.

If you are a €20–200M brand competing against someone larger, broad appeal is not a strategy. It is the absence of one — and it puts you in a fight you are structurally certain to lose, because the incumbent can outspend you on every dimension where you have chosen not to be specific.

The only viable path is to be more relevant to someone than the leader is, and accept being less relevant to everyone else. That is not a compromise. It is the mechanism. And that means making choices.

Who will think you are the most relevant choice for them?
The answer is never "everyone".

Which raises the harder question: relevant to whom?

Most companies answer this badly, because most segmentation is descriptive rather than causal. And making choices about who NOT to serve can feel painful. Age brackets, income bands, urban professionals, health-conscious consumers, mid-market B2B. These describe people. They do not explain choice. Two thirty-five-year-old urban professionals with identical incomes can want fundamentally different things from your category, judge value on different criteria, and choose completely differently. Grouping them together tells you nothing you can design against.

The method we use is Choice-Based Segmentation, and it starts from a different definition. A segment is not a group of similar people. It is a recurring pattern in how people choose.

Specifically: a segment exists where a shared desired outcome produces a repeatable choice pattern, held together by a recognisable identity pattern. The outcome clusters it. The choice pattern proves it. The identity makes it nameable and designable-for.

The test is unforgiving, and it should be. If you cannot say how this group chooses differently from the next one, you have not found a segment. You have found a description, or perhaps, a population.

That distinction disqualifies most of what passes for segmentation. A shared want is not a segment — plenty of people want the same thing and choose entirely differently. A shared circumstance is not a segment either; circumstance is usually the cause of an unmet outcome, discovered afterwards, not the basis for the cluster. And a segment that demands the same strategic answer as the one next to it is not a segment at all. It is decoration. It merges.

Demographics still have a role. They describe a segment, size it, and tell you where to find it. They may never define its choice logic. That is the inversion most segmentation work gets backwards.

Done properly, this produces two to five segments for a category, not twelve — and each one demands a genuinely different answer across positioning, identity, proposition and execution.

Then comes the commercial question, which is separate and should be kept separate: which of these are actually worth building toward? Attractiveness is a function of the value pool, the degree of unmet need, your ability to win, and your ability to reach them. A segment can be fascinating and commercially irrelevant. Another can be unglamorous and be the whole business.

The question worth taking into your next strategy session is not how to reach more people. It is this: which segments are most profitable for this brand specifically, and what would it look like to design the offering, the proposition and the brand around what those people actually want — rather than around the average of everyone we could theoretically serve?

That question tends to be uncomfortable, because the honest answer usually means giving something up. Narrowing the proposition. Accepting that some customers are not yours. Building something that some part of the market will actively reject.

Every one of those feels like a risk in the room. All of them are the price of being chosen rather than considered.

Finding "where to play" is half of strategy. The other half is knowing whether you are currently winning there — and most companies discover, when they finally measure it, that their brand was built for a customer who never existed.

The Relevance Diagnosis™ maps the segments in your market, defines who your brand is most built to win, and reads your positioning, identity, proposition and execution against what those customers actually choose on. Thirty days. Every gap named, evidenced, and priced. originalminds.co/diagnosis