Method
How we test
Every score comes from the same published rubric (version 2026.10-v1). Nobody can pay for a review, a better score or a faster place in line.
The process
- 01
Discover
We pull new launches from Show HN, Product Hunt, BetaList, DevHunt, Uneed, TinyLaunch, Fazier, Microlaunch, and direct submissions from makers.
- 02
Triage
Dead links, duplicates, spam and clones are dropped. Everything else joins the queue.
- 03
Automated checks
We load every app and check speed, HTTPS, privacy basics, security headers and exposed secret keys.
- 04
Hands-on test
An editor uses the app to complete its core job, start to finish, and scores it against the rubric.
- 05
Publish and re-test
Reviews go live with a dated score. Picks are re-tested monthly so we never recommend a dead app.
The Ship Score
| Dimension | Weight | What we check |
|---|---|---|
| Does it work? | 25% | Core flow completes, no broken buttons, no errors, data saves |
| Usefulness | 20% | Solves a real job better or cheaper than the incumbent |
| Experience | 15% | Time-to-value, clarity, mobile, accessibility |
| Performance | 10% | Load time, page weight, responsiveness |
| Trust & safety | 20% | HTTPS, privacy policy, no exposed keys, real maker |
| Staying power | 10% | Ship cadence, maker responsiveness, uptime |
What the badges mean
- Ship List Pick — the best app we've tested for a job. Requires a score of 85+.
- Also great — a strong alternative for a different kind of user.
- Worth watching — promising but early.
- Skip — we tested it and wouldn't recommend it today.
Security findings
If our checks find something serious, such as an exposed secret key, we tell the maker privately first and give them 14 days to fix it. We only observe what any visitor's browser receives; we never try to break in.