How to turn a startup idea into a validation plan

By Idea OS · Product and founder education

A startup validation plan is a written agreement about what you need to learn, how you will test it, and which result would change your decision. Start with the assumption most likely to invalidate the idea, rather than a long feature list.

1. Describe the customer and the current alternative

Name a reachable customer segment and the behavior you want to change. For a fictional scheduling product: independent fitness coaches who coordinate recurring appointments through messages. Check that this description matches real people before estimating a market.

2. Separate assumptions from evidence

StatementStatus at the startEvidence needed
Coaches spend too much time schedulingAssumptionObserve current work and ask about a recent incident
They will pay $19/monthProposed priceA real purchase decision with a clear offer
A booking link solves the problemSolution hypothesisUse during actual scheduling, including cancellations
A large market is reachableUnverifiedSourced customer counts and a practical acquisition test

Keep dates and sources with each observation. Do not turn a generated suggestion into an evidence claim.

3. Choose the smallest test that answers the question

If you do not know whether the problem matters, observe the current workflow. If the problem is clear but willingness to pay is unknown, offer a transparent paid pilot. If people pay but stop using the product, investigate repeated use before expanding acquisition.

For CoachSlot, first observe five coaches scheduling real appointments. Then consider a manually supported pilot, with its limitations explained. Five is a small discovery sample, not statistical proof of demand.

4. Decide what the result would mean before running the test

An illustrative threshold is: offer the pilot to five suitable coaches; continue investigating if three pay and use it weekly for four weeks. This is a proposed learning threshold for a fictional example, not a universal benchmark or a result that has occurred.

Write down the budget, time window, owner, and stop rule. If fewer coaches accept, record objections and review the segment, offer, and alternatives. Do not quietly change the success threshold after seeing results.

5. Record outcomes, including negative evidence

For each participant, record eligibility, the offer, acceptance or rejection, payment, actual usage, and what changed. Collect only the information you need and get permission before publishing identifiable quotes or outcomes. Report denominators: “three of five invited coaches paid” is clearer than “60% conversion” without context.

6. Make a bounded decision

Choose whether to continue, revise the segment or offer, repeat an inconclusive test, or stop. A small successful pilot justifies another learning step; it does not establish product-market fit.

Copy this validation-plan template

  • Customer and current alternative:
  • Riskiest assumption:
  • Evidence already available, with sources and dates:
  • What remains unknown:
  • Test and participant eligibility:
  • Offer and price, if applicable:
  • Proposed threshold and rationale:
  • Time limit, spending ceiling, and owner:
  • Observed outcome and denominator:
  • Decision and next question:

Use Idea OS to structure the review

The free hypothesis worksheet helps you write the customer, problem, offer, and test. Idea OS can then evaluate the assumptions and suggest next steps. You still conduct the experiment and verify the evidence.

See worked examples for a scheduling SaaS, a local services marketplace, and a consumer subscription. Each is fictional and illustrates a different risk.

Decide what to test before you build

Start with your customer, problem, and biggest unknown. Review the analysis, then collect evidence.

Evaluation requires a free account. Public worksheets do not.