What the work involves

You will spend most of your time on two activities: authoring marketing problems that a strong model should struggle with, and scoring model responses against structured rubrics. Task authoring means writing realistic prompts — a repositioning brief, a paid-social CAC problem, a portfolio architecture decision, a channel-mix reallocation under a fixed budget — along with a reference answer that shows your reasoning, not just your conclusion. Grading means reading model output closely and explaining, in writing, where the reasoning breaks: an invented benchmark, a growth tactic misapplied to a brand-building objective, a media plan that ignores frequency, a segmentation that sounds sophisticated but does not discriminate between buyers.

A second stream of work is rubric development. Marketing resists single correct answers, so much of the value you add is in defining what "good judgment" looks like in scoreable terms and calibrating with other SMEs so two reviewers grading the same response land in the same place. Expect written feedback on your own annotations early on and periodic calibration rounds throughout.

What the platform screens for

  • Depth over title. Eight-plus years at a recognized brand, agency, or holding company, with visible progression (Manager → Senior Manager → Director/VP). Screens probe for decisions you personally owned, including ones that did not work.
  • Prior rubric-based AI evaluation. This is stated as mandatory. Be ready to name the program or platform, describe the rubric dimensions, and explain a disagreement you had with a scoring guideline.
  • Written clarity. Feedback that a research team can act on is the deliverable; vague scoring notes are the most common reason candidates wash out.
  • Honest availability. Roughly 35 hours/week on weekday hours.

Logistics

Remote and largely asynchronous, though calibration sessions and SME syncs tend to land in US business hours. Cincinnatus LLC is the employer of record — W-2, payroll, benefits, and compliance — with placement inside the client lab's extended workforce. Pay in the $60–80/hr range reflects observed postings for this track and is not a guarantee; rates vary with seniority, specialization, and the specific workstream you are assigned.