One Email Beat Weeks of Building
Check whether real demand exists before you spend another week building, because the build has never been what proves an offer works. Payment is the strongest evidence; a real person asking is where it starts.
By James Schramko · Updated August 2026
You have an idea for the next thing. A product, a brand, a system, a piece of software. Nobody has paid a dollar for it yet, and you are already deep in the build. You are naming it, designing it, writing the copy and choosing the tech. The buyer only shows up once the thing is finished enough to show someone.
That order feels natural because building is the part you control. You cannot control a buyer. That is why you get everything else right first and hope the buyer turns up when it is ready. The build itself does not prove demand. Building can be part of a demand test. Finishing the build never proves anything. You can do every step of it well and still be building a room nobody wants to enter.
The Order Gets Reversed Constantly
Look at your own last few projects. Count how many started with the product before the buyer. Skip the market research slide and look at the actual first hour of work. Did you write the sales message, or did you open a design tool?
The founders who catch themselves in this loop usually know their marketing. They know they need customers at some point. What they miss is how much of the build is a way to put off the moment a real person can say no. A finished product is comfortable. A cold pitch to a stranger is not. Building buys time inside the comfortable part of the work.
What Building First Actually Costs
The real cost sits under the hours. It is the false confidence a finished build creates. Once something is built, it feels like progress. Progress feels like it deserves a launch. The launch then happens on a schedule set by the build, with no sign that any buyer was waiting.
If nobody buys, you blame the marketing, the price or the timing. Rarely the order you did things in. Weak demand gets mistaken for a marketing problem, and you go off and fix the wrong thing. You cannot tell which one you have until you test demand apart from the finished product. By the time the product is finished, that test has usually been skipped.
The Cheap Test That Tells You
A founder I coach was planning a special new offer. It was a real piece of work to design and price properly. He did not build the full mechanism first. He sent one email to his list asking who wanted in. No product built. No page live. That one email got a response rate of roughly five percent across sixty thousand recipients. The interested group filled a community of four hundred people within three days. The full offer got built after that number existed, and only then.
The email cost an afternoon. The other route was to build the whole mechanism first and find out afterwards whether anyone wanted it. That would have cost weeks and answered the same question far later and at a far higher price.
When It Is Not Even About Marketing
Sometimes the fastest version of this test skips the campaign altogether. A real person asks for the thing before you go looking for a market. A founder I coach builds AI agents. His first build was for someone he already knew, as a favor. That person was impressed enough to ask for a second build straight away, this time for his own company. The next day he sent a referral.
No positioning was tested. No landing page existed. The demand came first because a specific person with a specific problem asked for the thing. The founder never had to build a version aimed at nobody in particular. That is the fastest version of buyer before build. You already have one. Start there, instead of designing for a stranger who does not exist yet.
The Version Where Demand Was Real And Something Else Broke
Checking the buyer first does not remove every risk. It only removes the one risk you can check cheaply. A founder I coach ran a workshop and was paying about nineteen dollars for each registration. The registrations came in so strong that the team had to pause the campaign. One salesperson was off that week and another was not available, so more people wanted in than the sales side could handle.
That is a good problem to have, and it is still a problem. Registrations at about nineteen dollars each proved the buyer was real. They said nothing about whether the team could convert that many people at once. Test the buyer first. Just do not assume that answers every question the build still has to answer.
The Self-Diagnostic
Before your next build starts, answer these against the real project in front of you, not a project in general.
Has a specific person, not a persona, asked for this or shown real interest in paying for it? Could you run a version of this test today that costs an afternoon instead of a month? If the test came back with no interest, would you really kill the build? Or would you build anyway and blame the test later? What is the cheapest signal you could get before the next hour of build time? Have you gone and got it?
If you cannot point to a real demand signal, you are still guessing. The project plan does not change that.
Quick Reference
- A buyer is the only thing that answers the demand question.
- A cheap test beats a finished product every time. One email, one offer, one conversation. It answers the same question in a fraction of the time.
- One real buyer who already wants the thing beats a positioning exercise aimed at a stranger who does not exist yet.
- Proving demand does not prove you can deliver. Test both, in that order, before you scale either one.
- If you cannot name the signal that told you the demand was real, you have started building without a test.
Does a real buyer exist before the next build starts? It is one of the first questions I put to every founder who joins Mentor.