How to test a side project idea in a weekend before you build it

You can test a side project idea in one weekend without writing any product code. Spend Saturday talking to a handful of people about how they deal with the problem today, spend Sunday putting up a one-page landing page with a clear call to action, and judge the result only by what people commit to, not by what they say they like.

I think this matters even more when you build on the side. Your hours are limited, so the most expensive mistake is spending three months on something nobody needed. A weekend of testing is cheap insurance.

Saturday: problem interviews, not pitches

The best guide I know for this is The Mom Test by Rob Fitzpatrick. Its core idea is that people will lie to you to be nice, so you have to ask questions even your mom couldn’t lie about. The book boils it down to three rules:

• Talk about their life instead of your idea.
• Ask about specifics in the past instead of opinions about the future.
• Talk less and listen more.

In practice, that means you don’t open with “I’m building an app that does X, would you use it?” You ask things like:

• “When did this last happen to you?”
• “What did you do about it?”
• “What have you already tried, and what did it cost you?”
• “Is anyone else on your team dealing with this?”

If the person can’t remember the last time the problem came up, or they’ve never tried to solve it, that is useful information. A problem people don’t actively work around is usually not a problem they will pay to fix.

Five honest conversations are enough for a weekend. Message people you already know who fit the profile, a few online communities where they spend time, or former colleagues. Paul Graham’s essay “Do Things that Don’t Scale” makes the point that early users usually have to be recruited by hand, and the same goes for early interviews. Nobody will come to you.

Sunday: one page, one promise, one button

Once you’ve heard how people describe the problem in their own words, write a single landing page. It doesn’t need a logo or a pricing table. It needs three things:

1. A headline that names the problem in the words you heard on Saturday.
2. One sentence on what you’ll do about it, specific enough that the wrong people will leave.
3. One clear action: join a waitlist, book a 15-minute call, or pre-order.

Any no-code site builder is fine for this. The point is speed, not design.

The fake-door test

A fake-door test (also called a painted-door test) shows people a button or option for something that doesn’t exist yet, then measures how many click it. Amplitude describes it as a lean method that “validates demand before building.”

On your landing page, the fake door can be a “Get early access” or “Pre-order” button. When someone clicks, tell them right away that the product isn’t ready yet, explain that you’re checking demand, and offer to let them know when it is. Amplitude’s own guidance says to disclose the test immediately after the click and to keep sign-up optional. I agree. Tricking people is a bad way to start a relationship with your first users.

Share the page with the people you interviewed and in the same places you found them. A small amount of paid traffic can help if you want numbers from strangers, but it’s optional for a weekend.

What counts as a real signal

This is the part most people get wrong. The Mom Test separates compliments and vague future promises (“I would definitely use that”) from commitments. A commitment is when someone gives up something they value: their time, their reputation, or their money.

Here is how I’d rank what you collect, from weakest to strongest:

• Weak: “Cool idea.” “Let me know when it’s live.” Page views.
• Medium: An email on the waitlist. A click on the fake door.
• Strong: Someone books a follow-up call, introduces you to a colleague with the same problem, or agrees to try a rough version next week.
• Strongest: A pre-order, a deposit, or a letter of intent.

I don’t think there’s a universal number that means “go.” What I’d look for is a few strong signals from people who aren’t your friends. If everyone was polite and nobody committed to anything, that’s a no, or at least a “not like this.”

What to do on Monday

You’ll end the weekend in one of three places:

• Clear pull. People committed. Build the smallest thing that serves those specific people.
• Real problem, wrong solution. People described the pain in detail but didn’t bite on your offer. Rewrite the page around what they actually said and test again.
• No problem. Nobody could recall the problem or had tried to solve it. Let it go. You just saved months.

I wrote separately about protecting time for side projects with a job and a family, and why the “big weekend” can be a trap. This is the one weekend I’d make an exception for, because it decides whether the next hundred hours are worth spending at all.

Sources

• Rob Fitzpatrick, The Mom Test (summary by Yevgeniy Brikman)
• Paul Graham: Do Things that Don’t Scale
• Amplitude: Fake Door Testing