By Quick answer: How do I validate an app idea before building it?
- Interview — 10–15 people about past behaviour, not your idea.
- Check demand signals — search interest, competitors' reviews, communities complaining.
- Test action — landing page sign-ups, a fake-door button, or pre-orders.
- Deliver by hand — a concierge version for 3–5 customers.
- Decide — against pass/fail numbers you wrote down first.
Key takeaways
- Opinions are cheap; behaviour is evidence. Ask what people did, not what they would do.
- Competitors are good news — they prove demand. Their one-star reviews are your roadmap.
- The strongest signal is money or time committed before the product exists.
- Write the success criteria before you start, so you cannot move the goalposts.
- Validation is not a one-off; it continues through MVP, launch and every major feature.
Every app idea feels obvious to the person who had it. That is the problem. The feeling of certainty arrives long before the evidence, and it is the reason so many products launch to silence. CB Insights consistently lists the lack of market need among the top reasons startups fail3. Validation is how you find out — cheaply, quickly and before writing code — whether you are right.
The two-week validation sprint
| Days | Activity | Output |
|---|---|---|
| 1–2 | Write your assumptions and pass/fail criteria | A one-page validation plan |
| 2–7 | 10–15 problem interviews | Patterns in real behaviour and pain |
| 3–5 | Desk research: search interest, competitors, reviews | Evidence of demand and gaps |
| 6–10 | Demand test: landing page, fake door or pre-sales | Conversion numbers |
| 8–14 | Concierge: deliver the outcome by hand to 3–5 people | Proof the outcome is valuable |
Step 1: write down what must be true
List your riskiest assumptions and the evidence that would prove each wrong:
- The problem is real and frequent — at least 7 of 12 interviewees describe it happening monthly or more
- They are already trying to solve it — they spend time or money on workarounds
- They will act — 15% or more of landing-page visitors from the target audience sign up, or 5 people pre-pay
- You can reach them — you know where they gather and can get in front of them
Writing the numbers first keeps you honest. It is remarkably easy to reinterpret weak results as encouraging.
Step 2: problem interviews that tell the truth
The single most useful idea here comes from Rob Fitzpatrick's The Mom Test: people lie to be nice about your idea, but they cannot lie about their past1. So ask about what they did, not what they would do.
Ask
"Tell me about the last time this happened." "What did you do?" "What did that cost you?" "What have you tried?" "What do you pay for today?"
Avoid
"Would you use an app that…?" "Do you think this is a good idea?" "How much would you pay?" These invite polite, useless answers.
Our list of 50 user interview questions has more, by situation.
Step 3: read the demand signals
- Search interest — Google Trends shows whether people search for the problem and how that changes over time4.
- Competitors — their existence proves demand. Read their one- and two-star app-store reviews: that is a free list of unmet needs.
- Communities — Reddit threads, Facebook and WhatsApp groups, LinkedIn posts where your users complain.
Step 4: test whether people will act
Landing page
A one-page description of the outcome, with a sign-up or "join the waitlist" button. Drive a small, targeted audience to it and measure conversion.
Fake door
Add a button or plan for the feature inside an existing product or page; count clicks, then show an honest "coming soon" note.
Pre-sales
Offer early access at a price. Nothing validates like money. Dropbox famously used a demo video to fill its waiting list before the product was ready5.
Step 5: deliver it by hand
Before building software, deliver the outcome manually to three to five customers — a "concierge MVP". If you are building a meal-planning app, plan their meals over WhatsApp for two weeks. You will learn what they truly value, what they ignore, and whether they come back — all before writing code. Eric Ries's idea of the MVP is exactly this: the most validated learning for the least effort2.
Step 6: decide
Common validation mistakes
Asking friends
They love you. They are not your market.
Pitching instead of listening
If you talk more than a third of the interview, you learned less than you could have.
Counting compliments
"Great idea!" is not data. Commitments are.
Testing on the wrong audience
A thousand sign-ups from the wrong people prove nothing.
Key terms, defined
Validation — gathering evidence that a problem is real and people will act to solve it. Problem interview — a conversation about someone's past experience with a problem. Fake-door test — measuring clicks on a feature that does not exist yet. Concierge MVP — delivering the service manually before automating it. Pre-sale — payment before the product exists.
Frequently asked questions
How do I validate an app idea for free?
Interview 10–15 people who have the problem, read competitors' app-store reviews and community threads, check search interest with free tools such as Google Trends, and put up a simple landing page to measure sign-ups. Deliver the service manually to a few customers before building anything.
What questions should I ask in customer interviews?
Ask about specific past behaviour: when the problem last happened, what they did, what it cost them, what they have tried, and what they pay for today. Avoid asking whether they like your idea or would use it — people are polite and bad at predicting their own behaviour.
What is a fake-door test?
A fake-door test shows users a feature or product that does not exist yet — a button, a pricing option or a sign-up page — and measures how many try to use it. Those who click see an honest message that it is coming soon. It measures real intent cheaply.
How many people should I talk to to validate an idea?
Ten to fifteen good problem interviews usually reveal the main patterns. Keep going until answers repeat. For demand tests, decide in advance what conversion rate or number of pre-orders would count as a pass.
What if someone has already built my app idea?
That usually confirms demand. Study their reviews and pricing, find the users they serve badly, and decide whether you can do something clearly better for a specific group. Truly empty markets are rare and often empty for a reason.
Do I need an NDA to share my app idea?
Usually not. Ideas are rarely the scarce part — execution is — and most people you interview will not sign one. Share the problem openly to learn; protect code, data and brand through contracts and trademarks later.
Sources
- Rob Fitzpatrick — The Mom Test
- Eric Ries, Startup Lessons Learned — Minimum viable product: a guide (2009)
- CB Insights — Why startups fail: top reasons
- Google Trends
- TechCrunch — How Dropbox started as a minimal viable product (2011)
- Checked October 23, 2026. Figures that change — fees, salaries, prices — are reviewed every six months; if something is out of date, tell us.
Where to go from here
- What is an MVP? — the next step after validation
- 50 user interview questions — questions that uncover real problems
- Career guidesHow to validate a business idea in 2027: problem interviews, riskiest assumptions, landing page and pre-order tests, concierge MVPs, pricing signals, success criteria and deciding whether to buildHow to validate a business idea in 2027: problem interviews, riskiest assumptions, landing pages, pre-orders, concierge MVPs, pricing signals and decisions.
- How to make an appHow to build a SaaS product from scratchBuild a SaaS product from idea to paying customers: a painful problem, validation, a narrow MVP, auth, billing, multi-tenancy and your first 10 customers.
- Career guidesProduct discovery and validation skills employers want in 2027: customer interviews, opportunity solution trees, assumption mapping, prototypes and fake-door testsProduct discovery skills employers want in 2027: customer interviews, opportunity solution trees, assumption mapping, the four big risks and fast tests.
- How to make an appHow to make an app with zero coding experience (2027 guide)You can make an app without coding in 2027: validate the idea, design it, then build with no-code tools, AI app builders or a hired developer. Step by step.
- Templates & tools50 user interview questions that uncover real problems — grouped by goal, with the questions to avoid50 user interview questions for discovery and usability research, grouped by goal, with a simple script and the leading questions to avoid.
- Career guidesAI skills for founders in 2027: validating ideas, building AI MVPs, running lean, unit economics and defensibilityAI skills founders need in 2027: validating ideas, AI-built MVPs, lean operations, model costs and unit economics, defensibility and trust risks.
Get the ideas, not the noise.
Sign up with Google and we’ll send you short UX psychology tips by email — why people tap, stall and leave, and what to do about it. Unsubscribe anytime. Prefer something to keep? Get the free Laws of UX guide as a PDF.
Make sure the first screen makes sense.
A behavioral UX audit of your app or SaaS product: why users hesitate, stall and leave, and a prioritized fix list. Fixed price, 2–4 weeks.