How we rank
Every tool gets a score from 0–100. The formula is public, every input is shown on the tool page, and every change is logged.
| Component | Weight | How it's measured |
|---|---|---|
| Traction | 25 | Hacker News stories linking to the tool in the last 12 months (all tools), blended 50/50 with GitHub stars for open-source tools. Each signal is a percentile within the tool's use case so niche tools aren't crushed by giants. We do not use Product Hunt data. |
| Momentum | 15 | Change in traction signals over the last 30 days. Drives the weekly Movers & Fallers table. |
| Product quality | 25 | Live-checked on each tool site: docs/help, pricing page, API/developers, privacy policy, terms, changelog/release notes, security/trust page. Exact link-text or URL-path matches only; evidence URLs shown; a human spot-checks. |
| Value | 15 | Permanent free plan (a trial alone does not count), published pricing, and price versus the median for the use case. |
| Reliability / freshness | 10 | Weekly uptime ping, changelog or last-update date, valid SSL. Dead for more than 2 weekly checks = delisted. |
| Editor / community rating | 10 | 1–5 rating from the editor or users once there are enough votes; neutral 5/10 until then. |
Rules
- Paying never affects the score. Sponsored slots are labeled "Sponsored" and sit apart from the rankings.
- Each input is marked verified or provisional. Provisional inputs use a neutral midpoint (or an editor-entered value) until real data is collected.
- A tool needs at least 3 verified inputs to be fully Ranked; until then its score is labeled Provisional.
- Scores are recalculated weekly, and league tables show rank changes (▲▼).
- Small use cases: traction is ranked within the use case only when at least 8 tools in it have that signal; otherwise against all tools, so a leader of a tiny category can't get an inflated score.
- Blocked checks: if a site blocks our automated check, we carry forward its last successful check result for up to 14 days instead of resetting it, and say so on the tool page. After 14 days the input reverts to provisional.
- Publishing bar for new launches: tools from our daily collection (Show HN, GitHub, Hugging Face) must be a real product with a working site and meet at least one of: 200+ GitHub stars, 3+ Hacker News stories in 12 months, or 4/7+ on the quality checklist. Tools that don't yet qualify are listed as Rising and kept out of the overall top-20 until they do.
- Tool makers can dispute a score through our form; disputes go to a human review queue.
FAQ
Do paid placements change ToolStandings scores?
No. Paying never affects the score. Sponsored slots are labeled Sponsored and sit apart from the rankings.
What does a provisional score mean?
Each input is marked verified or provisional. A tool needs at least 3 verified inputs to be fully Ranked; until then its score is labeled Provisional.
How often are scores updated?
Scores are recalculated weekly, and league tables show rank changes (up/down).
How can a tool maker dispute a score?
Use the submit form on /submit/ and check the dispute box. Disputes go to a human review queue.
Methodology v0.1 (MVP). During launch most inputs are still provisional — we say so on every page rather than guess.