How to Validate Your App Idea Before You Build It

TLDR

• Validation is not opinion. Praise from friends means nothing. Real signal is people giving you time, an intro, or money before the product exists.
• Talk to 15-20 people first. Use The Mom Test: ask about their past and their current workarounds, never about your idea. Startups that spoke to 100+ potential customers had roughly 25% better odds of succeeding (createanmvp).
• Read demand that already exists in search volume, active communities, and competitors who charge money. If nobody searches for it and nobody sells it, that is a warning, not a gap.
• Run one cheap test this week: a landing page plus waitlist, $50-100 of ads, a concierge run where you deliver the service by hand, or a pre-sale. Total under $100.
• "Validated" means commitments, not clicks. Aim for a handful of people who paid a deposit or booked a call, not a spreadsheet of likes.

Most app ideas die not because the code was bad, but because nobody wanted the thing. The good news for non-technical founders is that you can find that out in a week or two, for the price of a nice dinner, without touching a builder or hiring anyone. This guide walks through the methods in the order you should run them, from cheapest and fastest to most convincing.

Step 1: Talk to people the right way

The cheapest test costs nothing and finds the most expensive mistakes early: talking to real people in your target market. The catch is that most founder conversations produce garbage data, because people lie to be nice. Ask someone "would you use an app that does X?" and they will say "sure, sounds great," then never think about it again.

Rob Fitzpatrick's book The Mom Test fixes this. The name comes from a simple idea: ask questions so specific that even your mum could not lie to you. Three rules do the work (The Mom Test, Sachin Rekhi's primer):

• Talk about their life, not your idea. Their opinion of your app is worthless. Their actual behaviour is gold.
• Ask about the past, not the future. "When did you last run into this problem?" beats "would you use X?" People are terrible at predicting future behaviour and great at describing what already happened.
• Listen more than you talk. The moment you pitch, they start being polite. Keep your idea out of the room.

Good questions sound like: what tools do you use for this now, what did the last time it went wrong cost you, and how are you working around it today. If they already built a messy spreadsheet or pay for a tool they complain about, that is a real problem worth solving. If they shrug and say it is not a big deal, believe them. Watch out for the three kinds of bad data Fitzpatrick warns about: compliments, vague statements about the future, and wishlists of features they will never pay for.

Aim for 15 to 20 conversations before you decide anything. You are not looking for a yes. You are looking for a problem that shows up again and again, described in the same words, that people already spend time or money trying to fix.

Step 2: Read the demand that already exists

Before you build anything, check whether the market is already telling you what it wants. Three free signals matter most.

Search volume. If people have your problem, many of them type it into Google. Google Trends shows whether interest is rising or fading, and a free browser add-on like Keywords Everywhere puts a real monthly search number next to those queries (Keywords Everywhere). A few hundred searches a month for a specific problem is a healthy sign. Zero searches usually means zero demand, not an untapped goldmine.

Communities. Go where your users already complain. Reddit, Facebook groups, and niche forums are full of people describing problems in their own unoptimised words, which surfaces the exact phrases and frustrations worth building around (ALM Corp on Reddit research). Search the subreddit for your problem. If you find threads where people ask for a fix and others reply "I gave up looking," you have found demand.

Competitors who charge money. New founders panic when they find competitors. They should celebrate. A rival charging a monthly fee is proof that people pay to solve this problem. Your job is not to be first, it is to be a better or more focused option for a specific group. The genuine red flag is the opposite: no search volume, no communities, and nobody selling anything. That silence usually means the problem is not painful enough to pay for.

Step 3: Run a cheap test this week

Conversations and research tell you a problem exists. A test tells you whether people will actually act. Pick one or two of these based on how far along you are. All of them fit inside a $100 budget.

Landing page plus waitlist. Build a single page that describes the product as if it already exists, with one clear promise and one button. Tools like Carrd or Framer do this in an afternoon. The email address is the signal. This is exactly what Dropbox did: Drew Houston posted a short demo video of a product that was barely built, and the waitlist jumped from around 5,000 to 75,000 people overnight (Thrive and Grow, Shortform). He validated demand before finishing the engineering.

Paid ads to that page. Warm friends will always click. Cold strangers will not lie. Spend $50-100 sending targeted traffic to your landing page and watch the signup rate. A conversion above roughly 10-15% from cold traffic is a strong sign the promise resonates (Capwave). Below a few percent, your promise or your audience is off.

Concierge. Before you automate anything, deliver the service by hand. If your app idea is "automatically match freelancers to briefs," go match five of them yourself over email first. You learn what actually matters, and you find out whether people value the result. The blunt test: if you cannot close three paying customers manually in a few weeks, more code is not the missing piece, demand is (RevenueCat).

Pre-sale. The strongest test of all. Ask people to pay now for access later, even a small deposit. Money is the only feedback that cannot be faked to spare your feelings. Ten people who each put down $20 tell you more than a thousand waitlist emails.

What "validated" actually means

Validation is not a feeling and it is not a like count. Fitzpatrick's rule is that real signal comes from commitments, where someone gives up something they value. There are three currencies:

• Time: they book a follow-up call, fill in a long survey, or test a rough prototype.
• Reputation: they introduce you to their boss or share your page with their team.
• Money: they pay a deposit, a pre-order, or commit budget.

A pipeline of concrete commitments is the only reliable signal that you are on the right track. Compliments are the opposite: they cost nothing, so they prove nothing.

Red flags to take seriously: everyone loves the idea but nobody will pay; you keep adding features because the core is not compelling on its own; the only enthusiastic people are other founders, not actual users; and you find yourself explaining why the low numbers do not count. If you are arguing with your own data, the data already won.

Your validation checklist

Run these in order. Stop and rethink the moment a step fails, rather than pushing forward on hope.

• 1. Write the problem in one sentence, not the solution. "Small landlords lose hours reconciling rent payments" beats "a rent app."
• 2. Do 15-20 Mom Test conversations. Ask about the past. Confirm the problem is real, frequent, and already costing time or money.
• 3. Check the demand signals: search volume, active communities, and at least one competitor charging money.
• 4. Build a one-page site with a single promise and a signup or pre-sale button.
• 5. Send $50-100 of cold traffic to it and measure the conversion rate.
• 6. Get a commitment: a deposit, a booked call, or a manual concierge sale. Aim for at least three.
• 7. Only then build. If steps 2 to 6 produced real commitments, you are building something people already want. If they did not, you just saved months and thousands of dollars.

None of this is glamorous, and that is the point. The founders who win are not the ones with the best idea in their head. They are the ones who found out cheaply, this week, whether anyone else cares. Do the boring validation first, and the building becomes the easy part.