Fake Door Test

A fake door test measures demand for a product by advertising it before it exists and counting how many people try to get it.

The name is literal. You put a door in the wall — a landing page, a pricing tier, a "Get started" button — and you record how many people reach for the handle. Behind it there is nothing yet. What you learn is not what people say they would do, but what they actually did when the option was in front of them.

How it works

The mechanics are always the same three pieces.

A door. Usually a landing page describing the product as though it already exists, with one clear action: join the waitlist, request access, pre-order.

Traffic that resembles your real market. This is the part most people get wrong. Sending the page to your own network measures how much your friends like you. The traffic has to come from somewhere your actual buyer already is — a relevant subreddit, a cold email list matching your ICP, a paid ad targeted at the segment you claim to serve.

A counted action. Not a pageview. Something that costs the visitor a little: an email address, a phone number, a card. The cost is what makes the signal worth reading.

A worked example

Buffer is the case everyone cites, and for good reason. Before writing the scheduling product, Joel Gascoigne put up a two-page site: the first explained what Buffer would do, the second showed pricing tiers. Clicking a plan led to a message saying the product wasn't ready yet, and a box to leave an email.

The pricing page is the important half. Plenty of people will join a free waitlist out of mild curiosity. Far fewer will click a paid tier. By separating the two, he learned not just that people wanted it, but that some of them had already decided it was worth money.

When to use it

A fake door test is the right instrument when your risk is demand, not feasibility. If you already know people want the thing and the open question is whether you can build it, this tells you nothing you need.

It is most useful before you have written any code, when the cost of being wrong is still measured in days.

When not to use it

Two situations call for something else.

If you cannot get honest traffic, the test cannot work. A fake door in front of an empty street measures nothing, and it is very easy to convince yourself that a low signal is a market verdict when it was really a distribution failure.

And there is the obvious ethical problem, which deserves more attention than it usually gets. You are showing people something that does not exist. That is defensible when you collect an email and tell them plainly that it is not built yet. It stops being defensible when you take money for a product you have no immediate plan to ship, or when you let someone believe they have bought something they have not. The honest version of this test is still a real test — it just says "coming soon" on the confirmation screen instead of pretending otherwise.

The related terms

A fake door test is one method inside problem validation — it tells you whether people want a solution, not whether they experience the problem in the first place. It pairs well with customer discovery, which explains why the people who clicked did so.

theLab runs fake door tests as its default mechanic: it generates the landing page, finds and contacts a real audience matching your ICP, and reports what they actually did. The verdict comes from signups and preorders, not from opinions.

theLab builds the landing page, finds real people who match your audience, runs the outreach, and returns a verdict from real signups. First experiment free, no card.