Data notes
Every chart here is drawn from data we collect ourselves, every day. The real time spans are short. Each chart shows its actual range. Short is short: we don't stretch or fill in gaps.
| Library | Actual range of the curve | Boundary |
|---|---|---|
| App Store | 2026-10-28 → 2026-11-08 (12 days) | 115 storefronts × 46 categories × Top Free / Top Paid / Top Grossing top 200, updated daily. Apple doesn't disclose downloads or revenue: rank and rating growth are relative signals. Apps that never charted aren't in the database. |
| Google Play | 2026-10-29 → 2026-11-07 (10 days) | 30 stores (no China store). Charted apps get their details pulled once a day; install counts are exact global totals. Detail pages refresh every other day, so a 0 increase on a given day doesn't mean it stopped. |
| GitHub | 2026-10-29 → 2026-11-08 (11 days) | The star counts are snapshotted once a day; points before 10-28 come from deeply and archived backfills, so they aren't continuous. Star counts can be gamed, and the database flags suspicious ones. |
| Indie product revenue (TrustMRR) | 2025-10 → 2026-11 (14 months) | Founders give TrustMRR a read-only key to their payment platform. Revenue is calculated by the payment platform, so it's trustworthy. Names, pricing, and channels are self-reported. Covers only about 12k products whose founders are willing to make revenue public. Month-by-month history goes back 12 months. |
| Meta Ad Library | 2026-11-01 → 2026-11-08 (8 days) | About 60 keywords × multiple countries of "currently running" ads. Advertisers on the first screen get pulled into the library, tracked every 2–3 days. 11-01 before it starts accumulating. Not finding an ad running ≠ no one is buying ads. |
| Product Hunt | Released 2013-11-24 → 2026-11-06 | Daily launches, vote counts, and rankings; starting at 10-29, we pull the launch-day top ~100 every hour. Product profiles are picked from here. |
Where these come from
- Opportunity Library: a scanner agent goes through the watchlist and every database each day at 06:00, maintaining a dozen-odd opportunities: status, evidence, window, who's building it, who it suits, confidence and what would kill it, plus daily changes. Each opportunity page is one of those entries, published as-is. The charts are drawn by looking up the App IDs, repos and domains cited in the evidence directly in the databases.
- Cutoff list: a candidate list every day. The rules are in each item's "lines it passed." These are candidates, not conclusions.
- Product dossiers: products picked from Product Hunt, one agent per dossier, written by checking the databases and reading the official site. Status and evidence strength are the dossier's own conclusions.
- One-line verdict: the line at the very top of the homepage, distilled by the model (DeepSeek) from the opportunity library's "Today's changes"; if that fails, it falls back to the first sentence of the original.
This edition: generated 2026-11-08 08:39,3991 on page. Opportunities: 94. Charted list: 7 days. Product profiles: 3,879. Curve entities: 2,382. Took 27 seconds.