How to Score a Startup Idea from 0 to 100

Most founders evaluate ideas by feel. A score forces honesty.

A scoring system makes you weigh the same evidence the same way for every idea, so a charming idea with no demand cannot beat a boring idea people are begging for. Here is a scoring system you can run yourself in a spreadsheet, built on the same principle we use at Demand Discovery: behavior counts, opinions do not.

The rule that makes scoring work

Score only what people do, never what they say. A survey "yes" is worth nothing. A public complaint is worth something. A payment to a bad workaround is worth a lot. If you let prompted opinions into the score, every idea scores well and the exercise is pointless.

The five components (100 points total)

1. Problem evidence (0 to 30)

Are people already talking about this problem, unprompted, in public?

• 0 to 10: you can find almost nothing. Nobody searches for it, nobody complains about it.
• 11 to 20: scattered mentions. The problem exists but people live with it quietly.
• 21 to 30: recurring, recent, emotional complaints in multiple places, plus steady or growing search volume for the problem language.

Collect real quotes with links. If you cannot paste five quotes from five different people, cap this component at 15.

2. Willingness to pay (0 to 25)

Is anyone already spending money or serious time on this problem?

• 0 to 8: no paid alternatives exist and nobody has built a workaround.
• 9 to 17: people use free tools or duct-tape workarounds and grumble about them.
• 18 to 25: people pay for imperfect alternatives today and complain about what they pay for. Complaints about a paid product are the strongest single signal in this entire system: they prove the money exists and show you the gap.

3. Market pull (0 to 20)

Is the problem growing, flat, or fading?

• 0 to 6: search interest declining, communities going quiet.
• 7 to 13: stable. The problem is durable but nothing is pushing new people into it.
• 14 to 20: a visible force is creating new sufferers: regulation, a platform change, a technology shift, a cost spike.

Score the trend of the problem, not the trend of your solution category.

4. Reachability (0 to 15)

Can you find and reach the people who have this problem, affordably?

• 0 to 5: buyers are scattered and unidentifiable, or locked behind gatekeepers.
• 6 to 10: they gather somewhere (communities, events, directories) but engagement is slow.
• 11 to 15: they self-identify publicly by complaining or asking for tools, so you can contact them directly today.

Test it: send ten messages to real people you found. Replies score points; silence does not.

5. Founder advantage (0 to 10)

Why you? Lived experience with the problem, distribution you already own, or skills that shortcut the build.

• 0 to 3: none. You found the idea, the idea did not find you.
• 4 to 7: relevant background or a warm network.
• 8 to 10: you have the problem yourself, or you already reach the buyers.

This is the smallest component on purpose. Founder fit tilts close calls; it never rescues an idea with no demand.

Reading the total

Below 40: kill or reframe. The demand evidence is not there. Killing an idea at this stage costs you a spreadsheet. Killing it after launch costs you a year.

40 to 65: not yet. Something real exists but a component is failing. Fix the weakest component with a targeted test (a landing page for willingness to pay, ten outreach messages for reachability), then rescore.

66 to 85: proceed. Move to product discovery: interviews, prototypes, an early sale, recruited from the people your evidence surfaced.

86 to 100: rare. Scores here usually mean strong incumbents saw it too. Ask why the gap is still open before celebrating.

Two rules keep the score honest. Set your kill threshold before you start scoring, not after you fall in love with the number. And score two ideas, not one: a score only means something in comparison, and your second-favorite idea will surprise you more often than you expect.

The fast way

Everything above is collectible by hand in one to two weeks per idea. Demand Discovery compiles it in minutes: every report shows the evidence behind your idea. Real people, quoted in their own words, asking for what you want to build, each quote tied to the source it came from, with a 0 to 100 score built on that evidence.

Start with a free Market Research report. Then $49, one time, no subscription, vets two ideas: two evidence reports, two landing pages, and one Agentic Launch to generate prospects for the winner.