Skip to main content

How to Validate an App Idea: A 2-Week Method

How to validate an app idea before designing or coding: problem interviews, review mining, search demand, smoke tests and pre-sales, with scripts and.

Insights10 min read1,829 words

This guide is for founders and indie hackers with an app idea and a limited budget who want evidence before committing months of design and development. You will get a two-week validation method, interview scripts, cheap demand tests, the signals that actually count, and a scorecard to decide whether to build, change direction or stop.

TL;DR: Validate the problem before the solution. Talk to 10 or more target users about their past behavior, mine competitor reviews for unmet needs, check search demand, then run a smoke test such as a landing page with a waitlist or a pre-sale. Decide your pass thresholds before you start, and treat money, time and specific commitments as evidence, while treating compliments as noise.

Why validation comes before design

CB Insights' widely cited analyses of startup post-mortems have repeatedly listed "no market need" among the top reasons startups fail. Building the wrong app is the most expensive mistake you can make, and design is not where it gets fixed: a beautiful app for a problem nobody has still fails.

Validation does not prove an idea will succeed. It reduces the biggest risks cheaply, in roughly this order:

RiskQuestionCheapest test
ProblemDo people actually have this problem, often and painfully?Problem interviews, review mining
DemandAre enough people looking for a solution?Search data, community research
WillingnessWill they sign up, switch or pay?Landing page, pre-sale, fake door
SolutionDoes your approach solve it in a way they use?Prototype tests, concierge MVP
ReachCan you get in front of them affordably?Small paid or community tests

Step 1: Write the riskiest assumptions down (Day 1)

Turn your idea into testable statements:

Target user: freelance designers in their first 3 years
Problem: they lose track of unpaid invoices and chase payments late
Current workaround: spreadsheets, calendar reminders, memory
Why now: more people freelancing, banks' invoicing tools are basic
Riskiest assumption: they care enough to pay for a dedicated app

Then set pass thresholds before testing, so you cannot rationalize weak results later. For example: "At least 6 of 10 interviewees describe chasing a late payment in the last month" or "At least 100 waitlist signups from 1,000 targeted visitors." Choose numbers that would genuinely change your decision.

If you are unsure who the target user is, sketch two or three candidate segments with the user persona generator and test the most promising first.

Step 2: Run problem interviews (Days 2–6)

Talk to at least 10 people in your target segment. Not friends who want to be supportive, and not people outside the segment.

The key principle, popularized by Rob Fitzpatrick's book The Mom Test: ask about specific past behavior, not opinions about the future. People are bad at predicting what they will do and polite about your idea.

Interview script

1. Tell me about the last time you [had the problem]. What happened?
2. How often does that happen?
3. What did you do about it? What did you try?
4. What does it cost you, in time, money or stress?
5. What tools or apps do you use for it today? What do you dislike about them?
6. Have you looked for or paid for a better solution? Why or why not?
7. Who else should I talk to about this?

Do not pitch your app until the end, if at all.

Signals that count vs noise

Strong signalWeak signal
Detailed story of a recent occurrence"Yeah, that sounds annoying"
They have built a workaround (spreadsheet, script)"I'd definitely use that"
They have paid for a partial solution"Great idea!"
They ask when they can try it"Let me know when it launches"
They introduce you to others with the problemGeneral enthusiasm with no specifics

A workaround is one of the best signals there is: someone who built a spreadsheet to solve the problem is already investing effort.

Step 3: Mine competitor reviews (Days 2–4, in parallel)

Existing competitors are a good sign: they prove demand. Your job is to find what they get wrong.

  1. •List 5 to 10 apps that solve the problem, directly or partly.
  2. •Read their 1- to 3-star reviews on the App Store and Google Play, plus Reddit threads and community forums.
  3. •Tag each complaint: missing feature, too complex, too expensive, bugs, wrong audience.
  4. •Count tags. Repeated complaints are your opening.
Complaint themeCountExample quoteOpportunity
Too complex for solo freelancers14"Built for agencies, not me"Simpler, single-user flow
No automatic reminders9"I still have to email clients myself"Auto-chasing
Expensive for low volume7"Pricing makes no sense for 3 invoices a month"Usage-based pricing

Step 4: Check search and community demand (Days 4–5)

Look for evidence that people are actively looking for a solution:

  • •Google Trends for the direction of interest over time.
  • •Keyword research tools for search volume on problem terms ("how to chase unpaid invoice") and solution terms ("invoice reminder app").
  • •App Store and Google Play search suggestions: type your core term and note the autocomplete phrases.
  • •Communities: count recent posts asking about the problem in relevant subreddits, Slack groups and forums.

Low search volume is not fatal for niche B2B apps, but zero activity anywhere is a warning.

Step 5: Run a smoke test (Days 6–12)

Now test whether people take action, not just talk.

Landing page with waitlist

Build a one-page site with the promise, three benefits, a visual of the app, and one call to action ("Join the waitlist" or "Get early access"). Drive targeted traffic from communities, newsletters or a small paid campaign. Our guide to building a landing page with AI gets you from idea to live URL quickly.

Make the visuals realistic. You can generate a first draft of the key app screens in floow to show on the page, so visitors react to something concrete rather than a vague description.

Stronger tests

TestHow it worksWhat it proves
WaitlistEmail signupInterest
Waitlist + surveyShort follow-up questions after signupInterest and segment fit
Fake doorA pricing page or "Buy" button that leads to "Coming soon"Willingness to start paying
Pre-saleDiscounted early access paid now, refundableWillingness to pay
Concierge MVPYou deliver the service manually for a few usersReal value and workflow
Wizard of OzA front end that looks automated, run manually behind the scenesWhether the solution works before you build it

Pre-sales are the strongest signal. Even a small number of people paying in advance tells you more than hundreds of signups. Be transparent: say it is early access, give a delivery estimate and offer refunds.

Step 6: Test the solution with a prototype (Days 10–14)

Once the problem and demand look real, test your proposed solution. Build a clickable prototype of the core flow only and put it in front of five target users. Watch where they hesitate and whether the flow delivers the outcome they described in interviews. Our app prototyping guide and the method in mobile app usability testing show how to do this in a few days.

Step 7: Decide with a scorecard

Score each area against the thresholds you set in Step 1.

AreaEvidencePass / weak / fail
Problem frequency and painInterview stories, workarounds
Gap in existing solutionsReview themes
DemandSearch and community activity
ActionWaitlist rate, fake door clicks
Willingness to payPre-sales, pricing reactions
Solution fitPrototype task success
ReachCost and effort to get visitors

Then choose:

  • •Build when problem, action and willingness to pay pass. Move to a focused MVP; see our MVP app design guide.
  • •Change direction when the problem is real but your segment, solution or pricing tested weak. Adjust one variable and retest.
  • •Stop when the problem itself fails. That result saved you months.

Common validation mistakes

  • •Asking "Would you use this?" Everyone says yes. Ask what they did last time.
  • •Interviewing friends and family. They are kind, not representative.
  • •Moving thresholds after the results. Decide pass criteria first.
  • •Counting traffic, not actions. Visits mean little; signups, clicks on "Buy" and payments matter.
  • •Validating with a generic audience. "Everyone with a phone" is not a segment.
  • •Skipping validation because a competitor exists. Competition is evidence of demand; find the gap.

FAQ

How long does it take to validate an app idea?

A focused validation sprint takes about one to three weeks: a few days of interviews and research, then one to two weeks of smoke tests. You can learn enough to make a confident decision well before spending money on development.

How many people should I interview?

Aim for at least 10 interviews within one well-defined segment. If answers are still very inconsistent after that, your segment is probably too broad; narrow it and interview again.

Can I validate an app idea without building anything?

Yes. Interviews, review mining, search research, landing pages, fake doors and pre-sales all work without code. A clickable prototype is the most you need before deciding to build.

What is the best signal that an app idea is valid?

Money and effort. People pre-paying, or already using a painful workaround, are much stronger signals than signups or positive comments.

Should I worry about someone stealing my idea?

Rarely. Execution matters far more than the idea, and the feedback you get from talking openly is worth much more than the small risk. Do keep genuinely proprietary technical details private.

Design it with floow

Once your idea passes validation, visualize it for your landing page or prototype test in floow. Try:

Design 3 iOS screens for an invoice-chasing app for solo freelancers: a dashboard showing overdue, due soon and paid invoices with totals; an invoice detail with an automatic reminder schedule; and a reminder email preview. Simple, friendly, realistic amounts and client names.

Design your mobile app with AI.

Generate pixel-perfect iOS and Android screens in seconds, then export to Figma or code.