Reading a Gap
A Gap is a recurring unmet need, assembled from many individual complaints. Every Gap carries a handful of numbers. None of them is a verdict, and each one deliberately ignores something. This page is how to read them like a skeptic.
Finding Gaps worth your time
Gap Explorer lists every Gap in the catalog. Sort by Opportunity Score for the most buildable, Friction for the angriest, Signal Volume for the loudest, or Newest for what just surfaced. Searching by idea switches to Best Match order.
The one filter worth turning on early: corroborated only. It hides Gaps whose evidence comes from a single product, which are the easiest ones to over-trust.
Opportunity Score: is it buildable?
A 0 to 100 judgment of how specific and buildable the Gap is, scored against a fixed rubric. It deliberately rewards sharpness over noise, so the loudest Gap is not automatically the highest scorer.
The rubric is strict by design: a typical Gap should not score well. If everything scored 80, the score would mean nothing.
Friction: how much does it hurt?
A 0 to 100 measure of how strongly people are expressing frustration inside this Gap. High Friction is real pain in real words, not mild preference. What it ignores: how many people feel it. A tiny group can be furious.
Products affected: is it one bad app?
How many distinct products the complaints come from. This is the number that separates a market gap from one product having a bad quarter. Thirty products generating the same complaint is a pattern. One product generating it thirty times might just be one broken release. Gaps backed by a single product are flagged.
There is no Momentum score here, and that is deliberate. We used to show one, comparing a Gap's last 30 days against the 30 before it. We removed it: we revisit a given source far less often than every 30 days, so the recent window was usually empty for reasons that had nothing to do with the market, and 43% of Gaps read exactly -100%. A number that reports our own collection schedule does not belong on a page about evidence. A growth measure returns when we can measure it honestly, and not before.
Signals and corroboration: how sure are we?
The signal count is how many individual pieces of evidence back this Gap. The product count next to it matters more: how many distinct products those Signals come from. Pain confirmed across many independent products is far stronger evidence than the same volume from one.
How to spot weak evidence
Three checks before you trust any Gap:
- Single-product evidence. If all Signals trace to one product, you may be looking at one company's bug backlog, not a market. We flag these with a warning.
- Stale evidence. Check the dates on the quotes. A pain that peaked two years ago may have since been solved.
- Read the actual quotes. The scores summarize; the quotes are the evidence. Every one links to its original source. If the quotes read differently than the Gap's name suggests, trust the quotes.
Going deeper on one Gap
Every Gap page includes a Theory of why the gap exists, its Missing Pieces (the specific capabilities people keep asking for, usually where a first version starts), related products, and the full evidence trail. Watch the Gap to get a weekly email when its evidence moves, or copy everything as Markdown for a spec or pitch.