How to test product-market fit before you launch

True product-market fit can only be measured after launch, because it requires real, engaged users experiencing your product. But the most expensive mistake in startups, building something nobody wants, happens before launch. So the goal pre-launch isn't to measure fit. It's to test demand cheaply enough that you never waste months building the wrong thing.

Here are five methods, ordered from weakest to strongest signal.

1. The landing-page test. Build a page describing the product as if it exists, drive a little traffic, and measure whether people click through to sign up. Cheap and fast, but a click is a weak signal, treat it as a directional read, not proof.
2. The waitlist. Add an email capture to the landing page. Now you're measuring willingness to give you something (an email and permission to follow up). A waitlist that fills without paid ads is real interest, and it gives you people to interview.
3. The fake door. Add a button for the feature or product inside an existing surface. When users click, show "coming soon" and capture intent. It measures demand in context, from people already in the relevant moment.
4. Pre-orders. Ask for money before the product exists. This is the strongest signal on the list, because payment is the one thing people won't do out of politeness. A handful of real pre-orders beats a thousand landing-page clicks.
5. The concierge MVP. Deliver the outcome manually, by hand, for a few real customers, before automating anything. You learn exactly what they value and whether they'll pay, with zero code.

The rule that separates real signals from noise

The reliability of a pre-launch test is proportional to what it costs the customer. An email is worth more than a click. A pre-order is worth more than an email. A concierge customer who keeps coming back is worth more than all of them. When you design a test, ask: what is the customer actually giving up here? If the answer is "nothing," the signal is weak.

This is the same discipline as the Mom Test: don't ask people what they'd hypothetically do. Create a small, real moment of commitment and watch what they actually do.

What pre-launch tests can and can't tell you

These methods de-risk the build. They tell you whether a problem is real and whether people will commit something to a solution. What they can't tell you is whether your actual product delivers, that requires shipping it and measuring problem-solution fit, then product-market fit with real users. Think of pre-launch testing as earning the right to build, and the Sean Ellis survey as confirming you built the right thing.