How to validate a startup idea: a 7-step checklist
Short answer: validating a startup idea means finding out, before you build, whether a specific customer has a painful problem, will pay a price that leaves you a profit, and can be reached cheaply. Do it in seven steps: one-sentence idea, real customer, painkiller test, competitors and substitutes, money maths, a cheap real-world test, then a pre-agreed kill / fix / test / go decision.
The 7-step idea validation checklist
1. Write the idea in one sentence
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 an idea yet, you have a topic.
2. Name the real customer
“Everyone” and “small businesses” are not customers. Write the job title or situation, where they spend time online, and what they already pay for. The test: could you message ten of them today?
3. Run the painkiller / vitamin / luxury test
| Type | What customers do | How it sells |
|---|---|---|
| Painkiller | Already searching, paying for workarounds, complaining unprompted | Bought urgently |
| Vitamin | Agree it is nice, plan to do it “later” | Postponed |
| Luxury | Like it, spend only when money is easy | Cut first |
4. Check what people do today
List direct competitors, indirect ones (spreadsheets, agencies, WhatsApp groups) and doing nothing. Strong competition usually proves demand; an empty market more often means no demand. Write one sentence on why a customer would switch to you.
5. Do the money maths
Estimate a realistic price, your monthly costs (tools, ads, support, payment fees) and the number of customers needed to reach your income goal. If the break-even customer count looks unreachable with the channel you have, the idea needs a different price, customer or channel.
6. Run the cheapest real test
Pick one: 10–20 customer conversations, a landing page with a clear offer and a sign-up or pre-order button, or a manual version of the service delivered by hand. Decide your success threshold before you start (for example, “5 of 20 people agree to pay”).
7. Decide: kill, fix, test or go all in
Kill when the problem is weak or the maths cannot work. Fix when the problem is real but price, customer or channel is wrong. Test when promising but unproven. Go all in only after real customers have paid or committed.
Common idea validation mistakes
- Asking friends and family, who are polite and not your customers.
- Treating “I would use that” as evidence. Money, time or a pre-order is evidence.
- Confusing a large market with a reachable customer.
- Ignoring hidden costs such as payment fees, support time and customer acquisition.
- Falling in love with the solution before confirming the problem.
Where an AI panel helps — and where it doesn't
A panel of AI models is useful for steps 1–5: it finds failure points, names likely competitors, proposes price ranges and flags assumptions you have not tested. Different models often disagree, and the disagreement shows you what to check first. It cannot replace step 6: only real customers can tell you whether they will pay. See how Delibra's panel works.
Frequently asked questions
How do I validate a startup idea quickly?
Write the idea in one sentence, name the exact customer, check what they do today, work out the price and break-even customer count, then run the cheapest real-world test you can (10 customer conversations, a landing page or a pre-sale). Decide in advance what result means kill, fix, test or go. A multi-model AI review can speed up the first three steps but cannot replace real customer evidence.
How do I know if my business idea is a painkiller or a vitamin?
Ask whether the customer is already spending time or money to solve the problem. If they are searching, paying for workarounds or complaining about it unprompted, it is a painkiller. If they say it sounds nice but would do it later, it is a vitamin. If they would cut it first when money is tight, it is a luxury.
Can AI validate a startup idea?
AI can pressure-test an idea: it surfaces failure points, likely competitors, pricing ranges and weak assumptions fast. It cannot know your market first-hand or predict success. Use AI opinions to decide what to test, then validate with real customers.
How many people should I talk to before building?
Around 10 to 20 conversations with people who match your target customer is a practical starting point. Stop when you keep hearing the same problem in the same words, or when it is clear nobody cares enough to pay.
Is there a free way to validate my startup idea with AI?
Yes. Delibra's Free Check runs a panel of three AI models on your idea and returns the verdict, the biggest failure risks, the real customer and a painkiller-vitamin-luxury rating. Paid plans add the money maths, competitor map, positioning, offers, ad hooks, a 7-day plan, red flags and a PDF.
Try it on your own idea — free. The Free Check runs a real panel of AI models on your idea and returns the verdict, the main ways it can fail, who the real customer is and whether it is a painkiller, vitamin or luxury. No card, no sign-up.
Run a free idea check