ChatBuildTest the idea before you build it
The most expensive thing you can build is the thing nobody wanted.
The most expensive thing a small business can build is the thing nobody wanted. Not because the build was costly, but because the months spent on it were.
Testing an idea first sounds like startup vocabulary. In practice it is one page, one week, and a decision you can defend.
What you are actually testing
Not whether people like the idea. Everyone likes ideas. You are testing whether someone will do a small inconvenient thing in exchange for it.
The inconvenient thing is the whole test. An email address, a deposit, a booked call, a reply to a message. Enthusiasm is free and means nothing; a tiny cost means everything.
The cheapest version that still tells you something
One page. What it is, who it is for, what it costs, and one action. No menu, no about section.
A real price. "From €X" beats "pricing TBC", because a price is what makes the interest honest. If you cannot name one, you are not ready to test.
A way to say yes that is slightly effortful: a form, a booking, a message.
A way to say why not. A single optional question on the confirmation — "what nearly stopped you?" — will teach you more than the numbers.
Then send people to it. Existing customers, your network, one small ad, a post where your audience already is. You need dozens of visits, not thousands.
Reading the result honestly
Set the bar before you launch, in writing. "If fewer than five people book a call in two weeks, I stop." Deciding afterwards is how people talk themselves into building anyway.
And separate the two failures. Nobody came means your description or your channel was wrong, and the idea is untested. People came and did nothing means the idea, the price or the audience is wrong. Only the second is an answer about the idea.
The honesty rule
Do not take money for something that does not exist, do not promise a date you cannot meet, and do not present a mock-up as a working product. A waiting list is honest. A checkout that charges for vapour is not, and it is the fastest way to lose the customers you were trying to find.
If you collect email addresses, say what you will use them for, and use them only for that.
Sometimes the test page is the product. More often than people expect, the one page that tested the idea turns out to be enough to start selling — which is why ChatBuild builds it from a description that takes about an hour of your time, rather than as a project. Start with one page.
What to do with a no
A no is the point. It cost you a week instead of six months, and it usually arrives with a reason attached from that optional question.
Most ideas that fail this test fail on price or on audience rather than on concept. Change one of those and run it again — that is a second week, and it is still cheaper than building.
Related: if the test says yes, how you describe it decides what gets built, and these are the four routes to building it.
Frequently asked
- How many visitors do I need for a useful test?
- Dozens, not thousands. You are looking for whether anyone takes the action at all, not for a statistically reliable rate, and for a small business the qualitative answers to "what nearly stopped you" usually teach more than the count.
- Can I take pre-orders for something I have not built?
- Only with complete honesty about what exists, what does not, and when it will. A waiting list or a refundable deposit with a clear explanation is fine; a normal-looking checkout for a product that does not exist is not, and it costs you the customers you were trying to find.
Sjors Hooijen Builds ChatBuild and writes about getting a business online without turning it into a project. Questions about this article? Email hi@chatbuild.com.