OpportunityMarket home

How to validate a startup idea before you write any code

The reason most validation produces false positives is that it asks about the future. "Would you use this?" is a question about a hypothetical, and people are generous about hypotheticals. Everything below asks about the past instead, where the answers are already fixed.

On this page

  1. 1What validation is, and what it is not
  2. 2The three questions
  3. 3A demand check you can finish in a weekend
  4. 4Signals that mean stop
  5. 5Signals that mean go

What validation is, and what it is not

Validation is establishing that a problem already costs identifiable people something measurable, and that they have already tried and failed to solve it. It is not a survey, a landing page with an email box, or a friendly conversation with someone who wants to encourage you. Those measure politeness and curiosity, both of which are abundant and free.

The three questions

Ask these of the evidence, and of any person you can reach. All three are about what has already happened.

  • When did you last hit this problem, and what did you do about it? A date and an action. Vagueness here means it is not a live problem.
  • What did you try before giving up, and what did it cost you? This produces your competitor list and your price ceiling in one answer.
  • What are you doing about it today? The honest answer is usually "nothing" or "a spreadsheet", and that is the thing you actually have to beat.

A demand check you can finish in a weekend

Pick the problem. Collect thirty public posts describing it, from at least three different places, all from the last ninety days. For each, record who wrote it, what they had tried, and whether money appears. Then count distinct people, count how many mention paying for anything, and split the thirty by date to see whether the second half is thicker than the first.

Three numbers, two days, and no code. If you cannot find thirty posts, that is the strongest result the exercise can give you, and it is worth more than the six months you did not spend.

Signals that mean stop

The evidence is one thread. Everyone describing the problem also names a free tool that solves it. Nobody mentions money anywhere. The posts are all older than a year. Every enthusiastic response comes from someone who would not be the buyer. Any one of these is enough to put the idea down and pick up the next one, which is the point of validating cheaply.

Signals that mean go

Strangers in unrelated places describe the same problem in different words. Several of them built a workaround, which means they paid in labour. At least one names a price they pay today and resents. The evidence is thicker in the last month than in the first. And the products that exist are visibly aimed at a different, larger customer than the people complaining.

Newly tracked, still early

Measured from public posts, updated continuously — every row links to its evidence

1
Lack of Drivers for Quick Order Acceptance
Delivery & Logistics Apps
44.1
2
Frequent Transaction Errors in Money Transfers
Banking & Payment Apps
25.2
3
Scentbird Subscription Service Issues
Customer Support
43.9
4
Frustrating User Experience with This App
Business Management Software
38.4
5
Offline Access for Mobile Apps
Mobile Apps
23.1
6
Frustration with Paycom Functionality
Business Management Software
20.3

Frequently asked

How do you validate a startup idea?
By finding evidence that the problem already costs specific people time or money, and that they have already tried and abandoned existing solutions — not by asking whether people would use your product.
Is a landing page enough to validate an idea?
It measures how well you write a headline. Email signups correlate with curiosity, not with willingness to pay, so treat the result as a weak signal at best.
How many customer interviews do I need?
Fewer than you think, if you ask about the past. Five conversations about what someone actually did last month beat fifty about what they might do next year.
Can an idea be validated without talking to anyone?
Partly. Public evidence answers whether the problem exists, how widely and in which direction. It cannot answer what someone will pay, which still takes a conversation.

Keep reading

Every guide

How to find, test and value an opportunity — the whole series.