How to validate a startup idea in two weeks
You can learn a lot about a startup idea in two weeks without writing code. This is a step-by-step plan with scripts and decision thresholds, followed by a worked example. For the broader theory, see the Idea validation guide.
Short answer: write the idea in one sentence, talk to 10 to 15 target customers, test a clear offer with a landing page, and decide using thresholds you set before you start.
No card and no sign-up for the free check. AI opinions, not predictions.
Days 1 to 2: write the bet
Use this shape: “[Specific customer] struggles with [specific problem], so they would pay [price] for [solution].” If you cannot fill every bracket, you do not have a testable idea yet. Then run a quick panel review with the startup idea validator to surface the weakest assumption. That becomes your first question for real people.
Days 3 to 8: talk to ten customers
Find people who match your customer, not friends. Ask about their past behaviour, not your idea. Good questions:
- “Tell me about the last time you dealt with [problem]. What happened?”
- “What did you try? What did it cost in money or time?”
- “What is the most annoying part?”
- “Have you paid for anything to solve this?”
- “Who else is involved in deciding?”
Avoid “Would you use this?” People say yes to be polite. Listen for specifics: a story, a number, a spreadsheet they built. The book The Mom Test by Rob Fitzpatrick is a well-known guide to this style of interview.
Days 9 to 12: test an offer
Build a one-page site with a clear headline, three benefits, a price and one button, such as “Join the pilot” or “Pre-order”. Send it to the people you interviewed and to the places your customer spends time. Count real actions: sign-ups with a reason attached, booked calls, deposits.
Days 13 to 14: decide with thresholds
Decide in advance what counts as success. Example thresholds:
| Signal | Go | Fix | Stop |
|---|---|---|---|
| Interviewees with a costly, frequent problem | 7 or more of 10 | 4 to 6 | Fewer than 4 |
| Agree to a paid pilot or pre-order | 3 or more | 1 to 2 | 0 |
| Can you reach 100 of them for under your budget? | Yes | Unclear | No |
These numbers are examples, not rules. Set your own, write them down, and honour them.
A worked example
Idea: a weekly meal planner for remote software engineers. Interviews: eight of ten say they skip lunch or order delivery, but only two have paid for a meal service, and most say they would try a free tool. Offer test: a landing page at a monthly price gets sign-ups but no pre-orders. Read: the problem is real but willingness to pay is weak. Decision: Fix. Test a narrower group that already pays, such as engineers with a fitness goal, or a corporate wellness buyer.
Illustrative example, not real data.
Frequently asked questions
How do I validate a startup idea with no money?
Use free AI reviews to find weak spots, interview 10 to 15 target customers, and run a one-page landing page test. None of this needs code or paid ads.
How many people should I talk to?
Ten to fifteen well-chosen target customers often reveals clear patterns. If answers vary widely, you may have more than one customer type and need to narrow down.
What is a good validation result?
People who gave time, data or money, not just compliments. Pre-orders, deposits and agreed pilots are stronger than sign-ups.
Should I build an MVP first?
Only after you have evidence of the problem and willingness to pay. Often a manual version delivered by hand is a better first test than software.
Keep reading
AskDelibra gives AI opinions and estimates, not predictions, guarantees, or financial, legal or investment advice. Always test with real customers.