Fake Door Test

Offering a feature that doesn't exist yet to see how many people try to use it.

Also known as: fake door test, fake door, painted door test

A fake door test puts a button, link or plan tier for something unbuilt — then measures who tries to open it. Clicks become the demand signal: a “Premium Export” button that shows “coming soon, leave your email” separates real interest from polite interview answers, at the cost of one afternoon.

add button → e.g. 4% of visitors click → collect emails → "not built yet, want early access?"
demand measured in clicks and contacts, not opinions

It works because the user behaves naturally — they do not know they are in a test. That honesty is also the ethical line: never take money for what does not exist, always reveal quickly, and follow up with the people who clicked (they are your warmest leads).

The classic mistakes:

Charging through the fake door. Collecting payment for vapor is fraud, not research. Interest and contacts only — money waits for pre-selling with a real delivery plan.
Clicks without context. A curious click is weak evidence; a click plus email plus a reply explaining the need is strong. Design the door to capture intent depth, not just counts.
Testing with your own traffic only. Your Twitter followers click out of loyalty. Run it on cold traffic or the signal measures fandom.
Leaving the door up. A fake button discovered months later by a real customer is a trust incident. Timebox every fake door and take it down on schedule.
Reading clicks as willingness to pay. Clicks measure curiosity; only money measures demand. Treat the door as a filter for what deserves a pricing test, not the pricing test itself.

Pair it: fake door for interest, concierge delivery for the clickers, pre-sale for the convinced — a ladder of escalating evidence before any code.