Pay index / History / CrowdGen by Appen
CrowdGen by Appen pay over time
CrowdGen by Appen has no published median yet: fewer than 5 distinct verifiable rates on every observed day.
Collecting since 2026-08-19: 5 observations so far. A series needs at least 14 observations across 14 days before we call it a trend. No line is drawn through fewer points than that would honestly support.
Every observation
| Day | Source | Median | Typical | Top | Distinct rates | Open roles |
|---|---|---|---|---|---|---|
| 2026-08-23 | snapshot | below floor | — | — | 3 | 32 |
| 2026-08-22 | snapshot | below floor | — | — | 3 | 32 |
| 2026-08-21 ▲ | snapshot | below floor | — | — | 3 | 32 |
| 2026-08-20 | snapshot | below floor | — | — | 3 | 36 |
| 2026-08-19 | reconstructed | below floor | — | — | 3 | 36 |
How to read this
- Snapshot: our own same-day reading of the platform's live feed.
- Reconstructed: rebuilt from our job-level ingestion records (first-seen, closed and rate-change logs) for the day before snapshots began. Rates are as our parser observed them that day; a rate it had not yet read contributes nothing.
- Archive: observed in a dated Internet Archive capture of the platform's own public listings page, parsed with the same rules as live data. The date shown is the capture date.
- ▲ 2026-08-21: Parser improvement: rates in listing descriptions became readable (all rate-change events logged that day are extraction artifacts, not market moves).
- Same rules as everywhere on the site: hourly only, one vote per distinct platform+rate band, midpoint percentiles, no median below 5 distinct rates, ceilings and per-task pay excluded. No point is interpolated; we never draw a line where nothing was observed.