What the work involves

Mercor builds high-fidelity simulations of a marketing org's real tool surface — email, CRM, project management, social media management, web and product analytics, AEO/SEO, ads, CMS and support — populated with realistic documents, dashboards, personas and deliberately planted problems. Your remit is the Creator and Influencer slice: creator discovery and evaluation, program decisions, seeding and campaign execution, usage rights and disclosure judgment, and share-of-voice contribution.

Day to day you will pull a task from work you have actually done — a creator whose engagement rate looks strong until you check comment quality, a whitelisting deal where the usage-rights window quietly expired mid-flight, a seeding wave whose attributed lift is entirely organic search — and turn it into a brief. That means specifying the tool states, the conflicting spreadsheets, the Slack thread with the wrong number in it, and then writing the reference answer plus the criteria that separate a genuinely strong response from a fluent, plausible, wrong one. You will also review AI agent attempts and grade them against that standard, which is where most of the writing effort lands.

What the platform screens for

  • First-hand program ownership. Screeners probe for creators you personally sourced, rates you personally negotiated, and measurement you personally defended — agency-adjacent or purely strategic exposure tends not to survive follow-up.
  • Disclosure and rights fluency. FTC endorsement guides, platform paid-partnership tooling, whitelisting and usage-rights terms, and what an ambiguous case actually turns on.
  • Evaluation judgment. Can you name the specific wrong-but-persuasive answer an agent will produce, and write a criterion that catches it?
  • Written reasoning. Much of the deliverable is prose explaining why a decision is correct. Terse or hedged reasoning does not carry.

Prior experience building AI training environments, RL environments or simulated case studies is strongly preferred, as is Rubric Academy or Rubric Bootcamp certification. Neither is stated as mandatory.

Logistics

Remote and asynchronous, contract, with observed rates in the $60–100/hr range — the band is what has been reported for this role family, not a guarantee. Expect a short multiple-choice knowledge screener specific to creator and influencer marketing before human review. Volume is typically part-time and elastic; contributors who can commit consistent weekly hours and turn around review batches within a few days get the most work.