Role-play

Pretend you are the talent ops lead

Pretend you are the talent ops lead and force the people system change to survive questions about leveling consistency, offer authority, interview fairness, and who owns the data when managers invent local shortcuts.

This seat sits between recruiting intent and the operating system that will be quoted in promotions and offers. It asks whether level definitions match how managers actually rate people, whether offer bands still bind under competitive pressure, and what happens when an interview loop skips a required step. A tidy handbook page does not answer those questions. The review needs the leveling rubric, the band tables, the loop design, and the named owner who can stop an out-of-policy offer.

Set the mandate

Give the seat the draft leveling framework, current offer bands, interview loop map, last quarter's exception list, and any open audit findings. Name the decision: clear the change, revise specific rules, or hold publication until a dependency is resolved. State which systems of truth the people tools must reconcile to.

The seat should distinguish sourced facts from management interpretation. It should flag a level that changed definition without a backfill plan, a band that managers routinely exceed, and any loop step whose interviewer pool is absent from the package. Recruiting owns pipeline; the talent ops lead owns system coherence and decision usefulness.

Ask for a people-ops question map

  • Which level definition moved, and what evidence supports the new boundary?
  • What changed since the last calibration, including any metric definition?
  • Which weak offer exception is easy to miss in the current reporting?
  • What interview skip still lacks an owner, and who remediates if it hits?
  • Who answers if a candidate challenges fairness or a manager challenges band authority?
  • Where could timing imply a promotion or raise the company did not intend to signal?

Require draft answers that stay inside the supplied facts. If the package cannot answer a predictable question, assign an owner or narrow the claim.

Run the room after the system freezes

Use Pingpong to simulate a short review where a hiring manager presses on one disputed level and a recruiter asks why the offer band moved from the previous update. Add compensation and interview loop responses from the same source pack. Note where speakers invent context, use conflicting denominators, or promise a follow-up nobody owns.

For leveling and offers, connect this seat to a leveling framework review, an offer comp band stress test, and an interview loop redesign. More review formats are listed in the war-game decisions hub.

Before the system goes out, have the talent ops lead annotate each material rule with its owner, definition, comparison period, and approved follow-up response.