Know If People Will Pay Before You Build: Fake-Door Test Checklist

You have an idea, no product, and no revenue, so the first test should cost little and answer one question before you incorporate or spend weeks building: will a stranger pay? A painted door is a button that promises a product you have not built, and you can put it in front of cold traffic. In hyped categories, a minimal sign-up prompt on a simple page can pull 8 to 15% of cold visitors even when the product has no credible sales path.

The floor separates interest from demand

Do not let a survey decide it. Stated interest and actual behavior often diverge by 2 to 3x in fake-door testing contexts. A landing page conversion benchmark is a stop signal. Use it to end a test, not to justify a build.

A validation test should feel small enough to abandon without guilt. A page that takes days to design has already started building. The page needs one promise, one audience, and one action. Do not add pricing, testimonials, or a roadmap. The visitor should understand what they are getting and what it costs them in time.

Use cold traffic. Friends and fans are a different audience. They will click because they like you, not because the offer is strong enough to buy. With a small budget, use a simple ad platform and send traffic to one page. You want a clean read, not a big launch.

The three tiers decide what the click means

A fake-door test becomes a willingness-to-pay test when the action asks for something real. Treat the action as the price of entry: the more it costs the visitor, the more it tells you. Notify-me is weak. It collects attention, not commitment. Choose it when you are testing a topic, not a buyer.

Email-plus-specifics sits in the middle, because the visitor has to say what they want and why they care. Ask for the problem, the budget, or the deadline. A visitor who will not type that is telling you the offer is probably too vague.

Card capture or prepay carries the strongest proof, because it asks for a payment signal before anything exists. A pre-order test can be as simple as a payment link for a future delivery. An awkward payment step is useful information.

88% of AI pilots fail to reach production, and most teams skipped demand validation before building.

The checklist turns a hunch into a go/no-go

Choose the tier before the page goes live, because the tier decides how you read the number. Run the test for 14 days on cold traffic, and treat anything under 2% conversion as noise. Require a call request as a second signal, then score the result against that tier before you bring it to friends or outside investors.

Do not change the page while the test is running. A necessary change means you note it in writing, then start the count again.

• Notify-me only: if it sits below the floor, stop.
• Email-plus-specifics: if it clears the floor, keep the list and ask for a call.
• Card capture on cold traffic: 5 to 15% is a real signal, and 15% or higher is a build-now case.

If a weak tier comes in below the floor, drop the idea or change the offer. If a strong tier clears the floor, build the smallest thing that can serve the first customers.

Write the result in one line first. A short note keeps the decision honest and makes the next step obvious for the next founder meeting.

Then do the next small thing: fix the offer, pick a different audience, or stop and say so plainly to one person.