Eighteen People Said Yes. Three People Bought.

You did it properly. You did not just guess.

You asked around — twenty or so colleagues, in person and in a group you belong to. Would you buy a course on this? Almost all of them said yes. Several said they had been looking for exactly that. One asked to be told when it was ready.

You built it. Three people bought, and one of them was a friend.

Nothing went wrong in the building. What went wrong was the instrument. Asking people about a purchase they have not made produces a polite answer about a hypothetical, and that is well understood. What is less obvious is what to do instead, because "talk to your market" is the advice everybody gives and it is what you already did.

So here is a different method. It takes four weeks, it involves asking almost nothing, and it looks for evidence of things people have already done.

Week one: collect complaints, ask nothing

Go where your profession talks — the groups, the forums, the association threads, the break room — and take notes. Do not participate. Do not float your idea.

Write down complaints verbatim. Their words, not your summary, because the phrasing is the part you will eventually need for your sales page, and you will unconsciously improve it if you paraphrase.

Point your attention at the places difficulty concentrates: skills the training never covered, work that is unpaid or underpaid, tasks that eat more hours than they should, things reputed to be hard, and situations where somebody clearly needed a mentor and did not have one. Those five categories will catch most of what is worth catching.

By the end of the week you will have a page of grievances. Most of them are noise. The next three weeks are about filtering.

Week two: find the workaround

This is the most useful single test I know and almost nobody applies it.

For every complaint on your list, ask: what have people built to cope with this?

Real problems generate workarounds. Somebody keeps a private spreadsheet because the system does not do the thing. Somebody has a document of copy-paste responses. Somebody always asks the same colleague. Somebody has a checklist they wrote themselves, taped inside a cupboard, that half the department has photographed.

A workaround is proof of pain that costs nothing to observe and cannot be faked by politeness. It means somebody spent their own time solving this because waiting was worse.

And the inverse holds: a complaint with no workaround anywhere is a grumble. It is real, it is annoying, and nobody is doing anything about it, which tells you exactly what they will do when offered a course about it.

Week three: ask what it cost, not what they would pay

Now you can talk to people, but change the question.

Do not ask whether they would buy something. Ask what the last occurrence cost them. How many hours. Which deadline slipped. What they had to redo. Whether it affected a review, a client, a role.

That question is about the past, so it has an answer, and the answer is a fact rather than a forecast. You will find out very quickly which of your candidate problems have a number attached to them and which produce a shrug.

You are also gathering the material that makes your eventual page persuasive, because a reader who recognizes their own cost in your description does not need to be convinced the problem exists.

Week four: follow the money that is already moving

Last filter, and the strictest.

Is anybody already spending on this? Look for existing courses, books, tools, consultants, an agency, a colleague being paid on the side, a conference session, an employer running internal training.

Money already in motion is the strongest signal there is. It proves the problem is worth paying to remove, it proves a budget exists, and it tells you the going rate. People fear finding competitors here; finding none is worse news than finding several, because it usually means you have located a problem nobody will fund.

Scoring what you have

You now have a short list, and each candidate is either carrying evidence or not.

A workaround exists. Somebody can name what the last occurrence cost. Money is already moving. Three signals, all of them about behaviour rather than opinion.

A candidate with all three is a business. A candidate with one is a maybe. A candidate with none is the thing eighteen people said yes to.

The trap worth naming

The problem you most want to solve is usually your own, from years ago, at the point where it hurt most.

Be careful with it. Your version of the problem may have been solved by something that has since changed, or may have been particular to where you worked, or may simply no longer be common. The observation weeks will tell you — if the complaints on your page do not include yours, that is the answer, and it is worth listening to even though it is the least welcome result the method can produce.

Build for the problem you observed, not the one you remember.