Validate a business idea before you build.

Before you start

Bring one idea and a way to reach people who might need it. This workflow suits a solo founder testing a small service or software concept. A pilot can test demand, but it cannot establish technical feasibility, safety, or regulatory readiness for a complex product.

Use four separate questions: Is the problem real? Can I reach the buyer? Will the buyer commit? Can I deliver at a workable cost? One positive interview does not answer all four.

1. Choose one customer and situation

Write who encounters the problem, when it happens, what they do today, and which assumption would make the idea fail if it were wrong.

“Tools for small businesses” leaves too much undefined. “Independent wedding photographers who chase missing client details after a booking” gives you a task to investigate and people to find. Treat the suspected cause as a question, not a fact.

2. Find the current alternatives

List the ways customers already handle the task, including doing it themselves. Record what each option includes, costs, and leaves unfinished.

Look at public offers, product documentation, and relevant discussions. Save the source and date for prices. A spreadsheet, an assistant, or accepting the inconvenience may compete with your idea even when no similar app exists.

For each alternative, write one question to verify with a customer. A missing feature on a competitor's page is not automatically a business opportunity. The customer may not need it, or may already have an adequate workaround.

The SBA's market-research guidance covers demand, existing choices, and pricing as inputs to a business decision.Read the SBA's business-planning guidance.

3. Ask about the last real occurrence

Interview people who recently faced the situation. Ask them to reconstruct what happened before showing your proposed solution.

Invite a small first group through relevant contacts or communities where research requests are welcome. Include people beyond supportive friends. Five conversations can be a manageable starting batch; that number is not a validation threshold or a representative sample.

1. Ask them to walk through the most recent occurrence, from trigger to outcome.
2. Find out which steps took effort, caused a delay, or needed another person.
3. Ask what they tried before, what they paid for, and why they kept or dropped it.
4. Identify who uses the solution, who chooses it, and who pays.
5. Ask what a worthwhile improvement would change in that specific workflow.

With permission, examine a redacted example such as a form or email sequence. Keep observed behavior, the customer's estimate, and your interpretation in separate notes. “Usually takes an hour” is an estimate until you have a basis for it.

For a conversation you can run, use the 20-minute outline, neutral follow-up questions, and completed evidence record in the interview lesson.Customer discovery interview questions and notes template.

Strategyzer explains why questions about past events provide different evidence from opinions about a proposed idea.Customer interviews: opinions and facts.

4. Test a commitment with a small offer

Offer the smallest useful result you can honestly deliver. State the scope, price, timing, customer inputs, and limits before asking for a commitment.

You may be able to deliver the result manually before building software. For Nina, that could mean setting up one intake form and three client emails in the photographer's existing tools. Call it a manual service; do not imply an automated product already exists.

If the customer needs approval before paying, record the actual next step and decision maker. Do not count a proposal, a waitlist signup, and a payment as the same result. Agree on what happens if you cannot deliver before taking money.

5. Decide what the result will change

Before running the test, set a time and cost limit, define the evidence that would justify the next step, and record what would make you revise or stop.

If the problem recurs but no one wants the offer, investigate scope, timing, trust, and price. If you cannot reach suitable people, improve recruitment before interpreting silence as weak demand. If evidence repeatedly contradicts the core assumption, reduce further spending and consider a different problem.

An accepted pilot opens a different question: can you deliver something useful at a sustainable cost? Keep that evidence separate from the initial sale.Deliver and review a paid pilot.

For more ways to separate assumptions and select an experiment, see Strategyzer's overview of business idea testing.Testing Business Ideas: method overview.

Your idea-validation worksheet

Copy these fields into a document. Use a new entry for each test, and keep customer details private. The example thresholds above are starting assumptions to adapt.

Customer and situation

Customer: ____. Trigger: ____. Task or problem: ____. Current alternative: ____.

Risky assumption

I believe: ____. Existing evidence: ____. What would contradict it: ____.

Research plan

Who qualifies: ____. How I can reach them: ____. Recent event to discuss: ____. Questions: ____.

Small test

Result offered: ____. Scope and exclusions: ____. Price: ____. Inputs and delivery date: ____. Commitment requested: ____.

Limits and decision rule

Time limit: ____. Spending limit: ____. Continue if: ____. Revise if: ____. Stop if: ____.

Evidence and next decision

Absolute counts and dates: ____. Observations: ____. Interpretations: ____. Still unknown: ____. Next decision: ____.

Plain text. Open it in any text editor or paste it into your notes.

Your next action

Write the customer, situation, and riskiest assumption now. Then identify one suitable person you can invite to discuss a recent example.

• Build a first-customer prospect list
• Work out a price for a small offer