The six questions
- Demand: is anyone actually looking for this?
- Market size: how big is it, really?
- History: who tried this before, and what happened?
- Competitors: how do they execute?
- Pricing: what will buyers pay?
- Verdict: go, pivot or not yet?
Why validate first
In CB Insights' widely quoted analysis of startup post-mortems, "no market need" was the reason founders gave most often (Fortune). Building is expensive. Checking is cheap. The aim is to find the reason not to build before you spend the money.
1. Demand: is anyone actually looking?
Look for people describing the problem in their own words, and paying to solve it today.
- What people search for, and how they phrase it.
- Forum threads, reviews and community posts that complain about current options.
- Job adverts: companies hiring someone to do this by hand have a real problem and a budget.
- What AI assistants answer when asked for help with the problem, and who they name.
2. Market size: how big is it, really?
Build the number from the bottom up: how many buyers have the problem, multiplied by what each would realistically pay per year. Show it as a range, and write down every assumption. One large number from a report is a headline, not an estimate.
3. History: who tried this before?
Almost every idea has been attempted. Find the earlier attempts, what they raised, and why they stopped. Then ask the question that matters most: what has changed since? A new technology, a rule change or a shift in buyer behaviour can turn a failed idea into a good one. If nothing has changed, the old reasons still apply.
4. Competitors: how do they execute?
A feature list isn't a competitor analysis. Look at what they do: how they position themselves, how they price, which channels they use, who they are hiring, and what their customers complain about. We cover the method in how to see how competitors actually execute.
5. Pricing: what will buyers pay?
Collect the prices buyers already pay for the closest alternatives. That gives you a band to work within before you speak to a single customer. See how to find what buyers will pay.
6. Verdict: go, pivot or not yet?
| Verdict | When it applies | What to do next |
|---|---|---|
| Go | Demand is visible, the price band supports the model, and you have a reason to win. | Run the cheapest test that risks real money or time from a buyer. |
| Pivot | The problem is real, but the segment, price or route to market is wrong. | Narrow to the segment with the strongest evidence and test again. |
| Not yet | Evidence is thin, or something that must be true isn't true yet. | Write down what would change the answer, and check back when it does. |
A good verdict also lists the three facts that would change it. That keeps the decision honest when new information arrives.
The checklist
- I can quote buyers describing this problem in their own words.
- I can name what they use today and what they pay for it.
- My market size is a range, with every assumption written down.
- I know who tried this before and what has changed since.
- I have looked at how five competitors actually operate.
- I have a price band built from real price points.
- I know the cheapest next test, and what result would stop me.
Questions people ask
How long does it take to validate a business idea?
It depends on the depth. A first pass on one question can take days. A full study covering demand, history, competitors and pricing takes longer. The point is to finish before you commit serious money.
Can an AI tool validate my idea for me?
AI tools are a fast first pass and useful for finding leads. They can also state things that aren't true. Treat every claim as something to check against a source before you rely on it.
What if the verdict is 'not yet'?
That is a useful result. Write down the facts that would change it, such as a price falling or a rule changing, and check again when they do.
Have an idea or a market to test?We answer these questions with sources and a clear verdict.
Scope a market study