How do you validate a startup idea without building it?

Confirm the problem through customer interviews focused on past behaviour, then run low-cost experiments: a landing page waitlist test, pre-sales, a concierge MVP delivered manually, or a smoke test measuring purchase intent. Pre-sales — people paying before the product exists — is the strongest signal.

What questions should I ask in customer validation interviews?

Ask about real past behaviour, not hypotheticals: what they do today, what it costs them, and the last time the problem hurt. Avoid 'would you use X?' questions, which invite polite but misleading answers.

All posts

StartupsJune 11, 2026 · 9 min read

How to Validate a Startup Idea Before You Build It

Most startups fail because they build something nobody wants. Here's a practical, low-cost playbook for validating demand before you write code.

To validate a startup idea, confirm that a specific group of people has a real, painful problem and will pay for your solution — before you build it. The goal of validation is to be wrong cheaply: to discover a flawed assumption in a week of conversations rather than after months of engineering.

Start with the problem, not the solution

Founders fall in love with solutions; customers care about problems. Before validating whether people want your product, validate that the problem exists, is painful, and is worth paying to solve. If the problem is weak, no amount of product polish will save you.

Talk to potential customers (properly)

Customer interviews are the highest-signal, lowest-cost validation tool — but only if you do them well. Ask about past behaviour, not hypothetical futures. 'Tell me about the last time you faced this' reveals truth; 'would you use a product that does X?' invites polite lies.

Ask what they do today and what it costs them in time or money
Dig into the last time the problem actually hurt
Avoid pitching — you're learning, not selling
Listen for whether they've already tried (and paid for) workarounds

Low-cost validation experiments

Landing page test: describe the offer, drive traffic, measure sign-ups for a waitlist
Pre-sales: the strongest signal — will people pay before it exists?
Concierge MVP: deliver the outcome manually for a few customers before automating
Smoke test: a 'buy' button that measures intent before anything is built

The strongest validation isn't a survey — it's someone handing you money before the product exists.

Define what success looks like in advance

Decide your pass/fail criteria before running an experiment, or you'll rationalize any result. 'If fewer than 20 of 200 visitors join the waitlist, we rethink the offer.' Pre-committing to the threshold keeps validation honest and protects you from confirmation bias.

When you're ready to build

Once you have evidence of a real problem and willingness to pay, scope the smallest product that tests your riskiest remaining assumption — a true MVP. Validation never fully ends; it just shifts from 'will anyone want this?' to 'how do we make it indispensable?'