What the work involves

You write the test, not the answer key alone. A typical task starts with a scenario grounded in something you have actually run — a Q3 reforecast after a demand miss, a variance bridge that has to be explained to a business partner who disputes the allocation, a capital request competing against two others under a fixed envelope. You build the world around it: the company profile, the driver structure, the data the model is given, the ambiguity that makes the task non-trivial. Then you draft the reference output (a forecast deck, a variance report, a business case, a model tab) and author a rubric that separates real FP&A judgment from spreadsheet-template output.

The rubric is usually the hardest part and the most scrutinized. Reviewers want criteria that fail a plausible-sounding wrong answer — a forecast that ignores seasonality baked into the driver set, a variance analysis that stops at price/volume without touching mix, a NPV that quietly assumes terminal growth above the discount rate. Generic scoring language does not survive review.

What the platform screens for

  • Ownership, not exposure. Screens probe whether you personally owned a budgeting cycle, a rolling forecast, or a capital allocation process — headcount, revenue scope, who you presented to.
  • Depth under follow-up. Expect two or three layers of questioning on a single answer: name the method, then justify the assumption, then defend it against a stated objection.
  • Reporting-basis fluency. US track answers should reflect US GAAP conventions; International track answers should be IFRS-consistent. Claiming both invites questions on where they diverge in practice.
  • Tooling specificity. Anaplan, Adaptive Insights, Planful, Tableau, Power BI, and large-company ERP environments — what you built in them, not that you saw them.
  • Writing. Rubric authorship is technical writing. Vague criteria are the most common rejection reason.

Logistics

Fully remote and asynchronous. Contributors typically commit somewhere in the 10–20 hour/week range, though this varies by project phase and is set per engagement rather than fixed. Work is contract-based and paid hourly; the $80–90/hr band reflects rates observed on this listing and is not a guarantee. Onboarding usually includes a calibration task or two before volume opens up, and continued work depends on rubric quality scores from reviewers.