How to fill your Mercor & micro1 profile
Your profile is not paperwork. On these platforms it is the input to a matching system that decides which projects you are shown, so a vague profile does not just look weak - it makes you invisible to the work you actually want.
There is also one mechanical rule worth knowing before you start, because it changes how you should sequence applications.
The reuse rule nobody mentions
According to Mercor's own documentation, each unique interview only needs to be completed once, and "results carry over to subsequent applications requiring the same interview." You also get up to three retakes across all applications for that interview, and there is a free practice interview that does not affect your opportunities.
| Rule | What it means for you |
|---|---|
| Interviews are reused | You do not re-interview for every role |
| Three retakes total | Do not burn them casually |
| Free practice available | Always take this first |
| 2–4 weeks to a decision | Apply elsewhere while you wait |
Practical sequence: take the free practice interview, fix what went badly, then take the real one once. Do not apply to six roles first and discover the format on the one you cared about.
Be specific about your domain
Mercor's guidance for new trainers is unusually direct: be specific about domains. Write "Clinical trial documentation editing" rather than "Health care."
The reason is mechanical. "Health care" matches everything and therefore ranks for nothing. A narrow, real specialism matches fewer projects but ranks near the top of those it matches - and those are the higher-paying ones.
| Too broad | Specific enough to match well |
|---|---|
| Health care | Clinical trial documentation editing, oncology |
| Legal | Commercial contract drafting, English law |
| Software | Distributed systems in Go; Kubernetes operators |
| Finance | IFRS 9 impairment modelling for retail banks |
| Writing | Technical documentation for developer APIs |
Claim transferable skills you actually have
Mercor's guidance also points at transferable skills - editing, research, tutoring, quality assurance - which map directly onto AI training work even if you have never called them that. A teacher who has written and graded exams has done rubric design. A QA lead has done structured evaluation.
The counterweight, from the same page: do not inflate. Calling yourself an "expert prompt engineer" without provable skill actively hurts acceptance, and copy-pasted bios or sloppy writing signal quality risk to reviewers.
“Claim narrow and true. The matching system rewards precision, and the interview punishes anything you cannot defend.”
micro1: pick one thing
micro1 asks you to choose one core skill or job type to be evaluated on. Take that literally. Picking your strongest single area produces a better certification result than spreading across three.
A note on the advertised average
Mercor's experts page advertises an average contracted rate of $120 an hour, while its own rate guide puts the AI-trainer average at about $31. Our own tracking of 178 hourly Mercor listings lands at $60–$80 typical.
All three can be true of different populations. Treat the high number as marketing, the low number as the broad market, and our figure as what is actually being advertised right now. Full detail in the pay report.
Profile checklist
- One clearly-named specialism, narrow enough to be real.
- Every skill something you could be questioned on.
- Transferable skills named explicitly (rubric design, structured evaluation, QA).
- Availability stated honestly - over-promising creates a delivery problem later.
- Free practice interview taken before the real one.
- A second platform applied to the same week.
Next: our Mercor interview guide and Zara interview guide.
Some outbound links to platforms are referral links.