Nobody Was Buying Marketing. I Went Back Through Ten Conversations to Be Sure.

We were writing a customer profile and it kept coming out wrong

Every version described the same person, and every version was useless.

We had all the material anyone could want — ten recorded conversations with real clients and prospects, transcribed, sitting in a folder. Not a survey. Not personas assembled from what we imagined. Ten hours of people describing their own situation in their own words while someone listened.

And the profile we kept producing read like a specification sheet. Specialty, revenue band, years in practice, what they had already tried. All of it accurate. None of it any use for deciding what to say to anyone.

So we stopped writing and went back through the transcripts looking for what they said when nobody had asked them a business question.

The sentence that showed up ten times out of ten

Every single one of them could name, precisely, the thing they were trying to get away from.

An employer. A group that owned the patient relationships they had built. A referral source that could switch off without warning. A private-equity owner who had changed the terms after the sale. One of them described a practice that simply stopped earning the moment they stopped physically working in it — a business that was, on inspection, a very demanding job with worse hours.

Ten conversations. Ten different specialties, cities and stages. Ten different versions of the same sentence.

Not one of them opened with a marketing problem. They opened with a dependency, and marketing was the exit they had settled on.

That reframed everything, and it is the line the whole profile now turns on: they were not buying marketing. They were buying the ability to stop depending on somebody else.

Why this is not a semantic distinction

I want to be careful here, because "sell the outcome, not the service" is exactly the kind of platitude that sounds like insight and changes nothing. This is more specific than that, and you can test it.

If someone is buying marketing, then more marketing is better. Cheaper marketing is better. Faster is better. You compete on volume, price and speed, and every competitor can beat you on at least one of them.

If someone is buying their way out of a dependency, almost none of that is true. What matters becomes: does this reduce what I am dependent on, and can I still be doing it in three years without you? Suddenly a cheaper option that keeps them dependent is not cheaper. Faster is irrelevant if it evaporates when the retainer stops. And the thing they most want to know about you is not your process — it is whether you will hand them something they own.

Two businesses selling identical services, priced identically, are not competing on the same axis at all depending on which of those they think is happening.

We had been answering the first question for two years. Every one of those ten people was asking the second.

The part we got wrong, and had to write into the profile as a hard exclusion

Here is the mistake that came out of the same review, and it is the more useful half.

There is a kind of practice that looks exactly like our best customer and is not one: established, busy, patients already flowing, adding a new service line alongside what they already do. On paper they are ideal. They have money, they have infrastructure, they are sophisticated buyers.

They also have no dependency they are trying to escape. Their patients arrive already. Their constraint is internal — capacity, conversion, operations — and none of it is solved by being more findable.

We had been selling to them anyway, because they said yes.

The profile now excludes them by name, in writing, at the top. Not as a nicety. Because an engagement with them fails in a specific and predictable way: we do good work, the numbers we control improve, and their business does not change, because we were never touching the thing that was actually binding. Everyone ends that relationship confused and slightly resentful, and they are right to be.

The uncomfortable version: saying yes to a customer whose constraint you cannot reach is not revenue, it is a deferred refund with extra steps.

What ten transcripts do that a hundred survey responses cannot

The thing that made this work was not sample size. Ten is a small number and I am not going to pretend otherwise.

What made it work is that nobody was answering a question we wrote. A survey can only return answers to what you already thought to ask, which means it structurally cannot tell you that your entire framing is wrong. Ten people talking freely will tell you that in the first twenty minutes, if you go back and listen for what they volunteered rather than what they replied.

Every one of those ten volunteered the dependency. Not one of them was asked about it.

That is the whole method and it costs nothing but attention. You almost certainly have the raw material already — recorded calls, sales conversations, onboarding sessions, support threads. The question to take back into them is not what did they say they wanted. It is what did they keep mentioning that nobody asked about.

Go find your own version of this

Pick five recorded conversations with people who bought from you. Read the transcripts, not your notes — your notes already contain your framing, which is the thing you are trying to escape.

Mark every sentence where they described something they were trying to get out of, away from, or free of. Then read only those sentences, in a row.

If they all say the same thing and it is not the thing on your website, you have just found the most valuable paragraph in your business, and it was already yours.

I wrote up what this looks like applied — who we work with, who we turn away and why, and what that means for what we charge — over on who this is not for. It is the most direct thing I have published, and it exists because of these ten conversations.