Validation field guide

How to Validate a Startup Idea Before You Build

A practical seven-day startup idea validation process with interview questions, evidence thresholds, and a go, pivot, or stop scorecard.

Validation means reducing a specific risk

Startup idea validation is not collecting compliments. It is finding credible evidence that a defined customer has a costly problem, is motivated to solve it now, and will make a meaningful commitment to your approach. The goal is not certainty. The goal is to replace an expensive assumption with enough evidence to make the next investment rational.

Begin by naming the assumption that could kill the business. An idea with obvious demand but impossible distribution needs a channel test. An idea in a familiar market with an unusual workflow needs a problem test. A product with strong usage and unclear economics needs a willingness-to-pay test. Test the riskiest assumption first, not the easiest one to discuss.

Write a falsifiable idea hypothesis
Customer: [specific role or situation]
Problem: struggles to [job] when [trigger]
Current alternative: uses [workaround] and dislikes [cost or failure]
Promise: our approach helps them [measurable outcome]
Commitment: they will [pay, pre-order, introduce us, or schedule a pilot]
Failure threshold: we will change direction if fewer than [number] of [number] qualified prospects commit

Use an evidence ladder, not a survey score

Different signals deserve different weights. A friend saying an idea sounds useful is weak evidence. A target buyer showing you their current spreadsheet is stronger. A signed pilot or payment is stronger still. Move up the ladder before you increase product scope.

SignalWhat it provesEvidence strength
Opinion or survey responseThe concept is understandableLow
Recent problem storyThe problem happens in real lifeMedium
Existing spend or workaroundThe problem already consumes money or timeMedium-high
Introduction or scheduled follow-upThe buyer will spend social capital or timeHigh
Deposit, pre-order, or signed pilotThe buyer will accept real riskVery high

Run a seven-day startup validation sprint

DayActionDeliverable
1Define the customer, problem, current alternative, and failure thresholdOne-page hypothesis
2Build a list of 30 people who match the segmentQualified outreach list
3Conduct five problem interviews without pitchingQuotes and repeated patterns
4Conduct five more interviews and rank alternativesProblem frequency and cost
5Show a lightweight offer, workflow, or concierge versionObjections and preferred outcome
6Ask for a concrete commitmentPilots, deposits, introductions, or scheduled trials
7Score the evidence and choose go, pivot, or stopWritten decision and next test

Recruit from where the problem naturally appears: role-specific communities, professional groups, former colleagues, review sites, support forums, or people paying for an adjacent product. Avoid filling the sample with people who want to help you but would never buy.

Ask for behavior, not predictions

A good interview reconstructs a recent event. It does not ask someone to imagine an ideal future. Stay with the last time the problem happened, what triggered it, how they handled it, what it cost, and why they chose that workaround.

  1. Tell me about the last time you tried to [job]. What triggered it?
  2. Walk me through what you did from the beginning.
  3. Where did the process slow down, fail, or require manual work?
  4. What have you tried already? What did it cost in time or money?
  5. Who else is involved in choosing or approving a solution?
  6. What happens if you do nothing for the next three months?
  7. If I could help with this outcome, what would a sensible next step look like?

Make the go, pivot, or stop decision explicit

Score every dimension from zero to two: zero means no evidence, one means mixed evidence, and two means repeated evidence from qualified prospects. Record the source beside each score so enthusiasm cannot quietly become evidence.

  • Problem frequency: the problem happened recently and repeatedly.
  • Problem severity: delay or failure creates a measurable consequence.
  • Current spend: buyers already spend money, labor, or political capital on it.
  • Reachability: you can name a repeatable channel to find similar buyers.
  • Commitment: prospects accept a concrete next step with real cost.
  • Founder advantage: you have access, insight, credibility, or execution speed others lack.
ScoreDecisionNext move
10–12GoBuild only the smallest product needed to test repeated use
6–9Pivot the testChange the segment, promise, channel, or commitment ask
0–5Stop for nowArchive the evidence and move to a stronger problem

Turn validation into the next controlled bet

If the idea passes, do not convert the whole vision into a backlog. Choose one narrow workflow and one behavior that would strengthen the evidence: a repeated task completed, a team invitation, a successful handoff, or a paid renewal. Your MVP should produce that learning with the least irreversible work.

Write the decision, the supporting evidence, what remains unknown, and the date you will review it. That record protects you from retelling the validation story later and gives collaborators a clear reason for what you build next.

Frequently asked questions

How many interviews are enough to validate a startup idea?

Ten well-qualified interviews are often enough to expose repeated patterns, but interviews alone do not validate demand. Seek a stronger commitment such as a pilot, deposit, paid concierge test, introduction, or scheduled implementation step.

Can I validate a startup idea without building a product?

Yes. Problem interviews, a manual concierge service, a clickable workflow, a landing-page offer, and a paid pilot can test the customer, urgency, promise, and willingness to pay before software exists.

What is the biggest startup validation mistake?

The most common mistake is treating positive opinions as buying intent. Validation should measure recent behavior and a concrete commitment from a specific customer segment.