How we score
One number you can rank by, and a second that tells you how much to trust it.
The five signals
| Signal | Weight | What it measures |
|---|---|---|
| Pain | 30% | How severe and frequent the problem is, and whether people already spend money or time on it. It leads because it is the best single predictor of a product people pay for. |
| Why now | 25% | What recently changed in technology, regulation or behavior that makes this newly possible or urgent. |
| Revenue potential | 20% | A realistic ceiling for a small team, from price points and the number of reachable buyers. |
| Open market | 15% | How much room incumbents leave. 100 means wide open. |
| Ease to build and sell | 10% | Whether one to three people can ship it and reach buyers. Distribution counts as much as code. |
The evidence multiplier
After weighting, the total is multiplied by a factor between 0.70 and 1.00 based on the Evidence score: how much real-world proof supports the signals, such as paying comparables and documented demand. An idea with an Evidence score of 50 is multiplied by 0.85, so it can’t outrank a well-proven idea on enthusiasm alone.
Verdicts
Build it at 72 and above, refine first from 55 to 71, and pass for now below 55. A verdict is a starting point for customer conversations, not a substitute for them.
Data tiers
Payment-verified revenue read directly from a payment provider. Self-reported figures founders publish themselves, weighted lower. Modeled estimates, mostly for mobile apps, shown for context only.
The revenue board in this build is a fictional demo dataset. Comparables marked “named by AI · verify” come from the model’s general knowledge and should be checked.
Limits
Scores are AI estimates produced from your description and our dataset. They are for educational purposes only and are not financial, legal or investment advice.