Help me validate this business idea: [BUSINESS IDEA]. The target customer is [TARGET CUSTOMER], the problem evidence I have is [CURRENT EVIDENCE], and my resources and constraints are [RESOURCES AND CONSTRAINTS]. My alternative solutions or competitors are [ALTERNATIVES]. Do not assume the idea is viable and do not invent market size, customer demand, pricing evidence, competitor features, interview quotes, or unit economics. Treat my statements as hypotheses unless I provide supporting evidence. Produce a validation brief with these sections: 1. Problem definition: customer, job to be done, current workaround, trigger moment, and cost of inaction. 2. Assumption map: list the 8-12 assumptions across problem severity, customer segment, willingness to change, distribution, pricing, delivery, retention, and economics. Rate each by importance and uncertainty.
3. Evidence review: separate what my inputs support from what remains unknown.
4. Customer discovery plan: identify 10-15 interview targets, recruitment routes, and 8 non-leading questions. Do not write questions that pitch the solution or ask whether people “would use” it.
5. Test plan: propose the smallest ethical tests that can reduce the three riskiest assumptions, with sample size guidance expressed as a learning threshold rather than statistical proof, success and failure signals, cost, duration, and stopping rule.
6. Alternatives comparison: explain why customers might retain [ALTERNATIVES], switch, or do nothing.
7. Recommendation: choose one next action—interview, landing-page test, concierge test, prototype test, pilot, revise, or stop—and explain what result would change the recommendation.
Flag areas that need legal, financial, safety, privacy, or regulated-industry review before customer testing. Do not advise deceptive testing, fake scarcity, or collecting sensitive information without a proper process.
Before answering, check that each test measures behavior or a credible commitment rather than stated enthusiasm, and that the recommended action matches the highest-risk assumption. Ask up to three clarifying questions only if a required input is missing.Fill in
| Placeholder | What to enter | Example |
|---|---|---|
| [BUSINESS IDEA] | Describe the proposed product or service and the problem it aims to solve. | A service that turns independent dental practices’ overdue-treatment lists into personalized follow-up campaigns. |
| [TARGET CUSTOMER] | Define the initial customer segment, their role, and buying context. | Practice managers at 3-10 dentist clinics using Dentrix and employing no dedicated marketer. |
| [CURRENT EVIDENCE] | List interviews, observations, sales data, research, or experience that support the idea. | Five managers said follow-up is inconsistent; two shared that staff spend several hours weekly calling patients. |
| [RESOURCES AND CONSTRAINTS] | State your budget, time, team, technical capability, and relevant limits. | One founder, $3,000, eight weeks, no access to patient data until a compliant process is in place. |
| [ALTERNATIVES] | List current customer workarounds, direct competitors, and the option to do nothing. | Front-desk calls, existing practice-management reminders, local marketing agencies, and no outreach. |
How to use
- Describe the customer narrowly enough that you could name ten people who fit the segment.
- Paste only evidence you can trace to interviews, observed behavior, data, or credible research.
- Run the first test before building broad features, and record both disconfirming and confirming results.
- Send this follow-up: “Write a 30-minute discovery interview guide for the highest-risk assumption and a coding sheet for the answers.”
Variations
Interview guide
Use this when you need to learn about the problem before showing a solution.
Create a customer-discovery interview guide for [TARGET SEGMENT] about [PROBLEM AREA], based on [CURRENT HYPOTHESES]. Produce a recruiting screener, a 30-minute interview flow, eight open questions, neutral follow-ups, and a note-coding template. Ask about recent specific incidents, current workflow, alternatives, cost, decision process, and prior attempts to solve the problem. Do not ask respondents to design the product, predict hypothetical use, or react to a pitch. Check that each question can reveal behavior from the past rather than preference. Ask up to three questions only if a required input is missing.
Landing-page test
Use this when you need to test message and demand before building.
Design a landing-page validation test for [OFFER] aimed at [TARGET CUSTOMER], using [CHANNEL AND BUDGET] and [CURRENT EVIDENCE]. Produce page sections, three value-proposition variants, one credible call to action, qualification questions, event tracking, traffic plan, and decision thresholds. Define what a conversion means and what it cannot prove. Avoid unsupported claims, fake testimonials, fake availability, or collecting sensitive data unnecessarily. Include a follow-up path for qualified leads. Check that the test measures a meaningful commitment, not only page views. Ask up to three questions only if essential detail is missing.
Pricing test
Use this when the problem is credible but willingness to pay is unknown.
Plan a willingness-to-pay test for [SOLUTION] with [TARGET BUYER], based on [PROBLEM EVIDENCE] and [DELIVERY COSTS OR CONSTRAINTS]. Return price hypotheses, interview questions about current spend and alternatives, a proposal or concierge-test structure, acceptable commitments, and interpretation rules. Separate willingness to discuss price from a real buying signal such as a paid pilot, signed letter of intent, or budget-owner introduction. Do not invent benchmarks or claim price validity from a small sample. Check that the test includes a clear reason to stop or revise. Ask up to three questions only if required inputs are missing.
Tips
- Start with a narrow customer segment and a recent painful workflow; a broad audience makes every interview signal difficult to interpret.
- Prioritize evidence of existing behavior such as budget, workaround effort, access to a buyer, or a paid pilot over compliments about the concept.
- Do not make a landing-page conversion rate mean more than it does; traffic source, message, price, and call to action can each explain the result.
- In regulated or sensitive domains, validate the problem and workflow first, then obtain appropriate legal, privacy, security, or clinical review before handling real data or making claims.
FAQ
Can ChatGPT validate a business idea by itself?
No. It can help design tests and challenge assumptions, but validation requires evidence from real customers, behavior, and operational constraints.
How many customer interviews should I conduct?
Conduct enough interviews to test a focused hypothesis and watch for repeated patterns, then run a behavioral test. The right number depends on how homogeneous the segment and problem are.
Should I build an MVP before validation?
Usually test the riskiest assumption first with interviews, a concierge service, prototype, or landing page. Build only what is necessary to learn whether customers will act.