SignalGigs original data
2,005 postings did not mean 2,005 different roles
One source's 2,005 posting instances reduced to 22 reviewed role groups. Headlines and location variants made the opportunity pool look much larger than the underlying set of tasks.
Case study observed October 30, 2026. Published October 31, 2026. This historical audit stays dated when other reports refresh.
2,005
Posting instances
22
Reviewed role groups
49
Headlines for one role
What a posting count can hide
The largest group contained 674 posting instances under 49 headlines. A repeated headline can reach a new location or audience without creating a new type of work. These counts cannot tell us how many people a role will hire, or how many paid hours it will supply.
| Representative role title | Posting instances | Distinct headlines |
|---|---|---|
| $110 for One Visit: Paid Research Study, No Preparation Needed | 674 | 49 |
| Licensed Physician (MD/DO) - Freelance AI Trainer | 530 | 16 |
| Senior CFD Engineer (OpenFOAM / ANSYS Fluent) - Freelance AI Trainer | 232 | 4 |
| Freelance Mechanical CAD&Simulation Engineer - AI Trainer | 231 | 4 |
| Bezahlte Studie in New York - Deutsch-Muttersprachler:innen aus Österreich | 34 | 28 |
| Earn up to $1,600 - 3-Day Arabic Language Study in NYC | 34 | 32 |
| $100 Half-Day Paid Study - Home & DIY Tasks, Los Angeles | 33 | 33 |
| Bilingual Italian-English Participant - Paid UX Study, New York | 18 | 17 |
| Native Polish Speakers in NYC: Paid In-Person Research Study (Up to $1,600) | 17 | 17 |
| European Portuguese Speaker Study Participant - NYC, $1,200-1,600 | 17 | 17 |
| Earn up to $1,600 - 3-Day Swiss Italian Language Study in NYC | 17 | 17 |
| Bilingual Swiss German-English Participant - Paid UX Study, New York | 17 | 17 |
| Native Hebrew Speakers in NYC: Paid In-Person Research Study (Up to $1,600) | 17 | 17 |
| Earn up to $1,600 - 3-Day Korean Language Study in NYC | 17 | 17 |
| Bilingual Japanese-English Participant - Paid UX Study, New York | 17 | 17 |
| Native Swedish Speakers in NYC: Paid In-Person Research Study (Up to $1,600) | 17 | 17 |
| Native Danish Speakers in NYC: Paid In-Person Research Study (Up to $1,600) | 17 | 17 |
| Native Finnish Speakers in NYC: Paid In-Person Research Study (Up to $1,600) | 17 | 17 |
| Earn up to $1,600 - 3-Day French Language Study in NYC | 16 | 16 |
| Dutch Speakers Wanted - Paid Smartphone Study, NYC | 16 | 16 |
| Earn up to $1,600 - 3-Day Flemish Language Study in NYC | 16 | 16 |
| Native Dutch Speakers from Belgium in NYC: Paid In-Person Research Study | 1 | 1 |
| Total in reviewed sample | 2,005 | 402 |
A better way to shortlist
Compare the underlying tasks, applicant conditions and destination before counting a new headline as a new opportunity. Keep language and location conditions visible: two variants may share work while serving different eligible applicants. A headline count alone cannot measure choice.
Method and limits
A retained October 30 review grouped source posting instances by the underlying role and mapped redundant saved records to one representative. Counts below are recomputed from that reviewed mapping, not inferred from matching titles alone.
The public JSON contains all 22 group totals and 381 reviewed duplicate-to-representative mappings from the saved audit. Summing the group posting counts reproduces 2,005; summing their distinct-headline counts gives 402. The maximum headline count is 49.
The audit concerns one source on one date. We do not extrapolate its repetition rate to other sources. Descriptive role titles are retained to make the grouping reviewable; pay appearing inside those historical titles is not a current offer.
- A single-source case study on October 30, 2026; not a current market-wide duplicate rate.
- The mapping is an editorial judgment. Location or language variants may legitimately target different applicants.
- A role group is not a count of vacancies, hires, projects or paid hours.
- The retained audit contains group totals and reviewed saved-record mappings, not every original posting instance. Totals can be checked; the full source collection cannot be independently reconstructed from this export.
Method version 2026-10-31.1. Downloaded JSON includes the input fingerprint for reproducibility.