| name | performance-review-packet |
| description | Use when preparing or running a performance review for an employee — assembles the rubric, goals, and evidence into one packet so a rating is defensible, calibrated, and turns into a real next step. |
Performance Review Packet
A playbook for turning a review from a recency-driven opinion into an evidence-backed, calibrated judgment that produces growth. The failure mode it prevents: a rating built from the last thing the manager remembers, graded on a scale only that manager uses, that ends in a number and no next step.
Set the frame
- Start from the rubric: the competencies and the defined expectations at this level. If the employee doesn't have goals on record, that's the first gap to fix — you can't grade against a target that was never set.
- State the period and the level being graded to, so "meets" has a fixed meaning.
Assemble the evidence
- Gather what actually happened: goals set and their outcomes, work delivered, and feedback from peers and stakeholders. Keep observed facts separate from conclusions.
- Draft a proposed rating per competency, each backed by specific evidence — not a general impression. A rating without evidence is an opinion wearing a number.
- Flag the thin spots: competencies where you don't have enough evidence to grade honestly, so calibration can probe them.
Calibrate
- Compare the ratings against peers at the same level and across managers, so a "meets" here equals a "meets" elsewhere. Surface the outliers — the unusually high, the unusually low, the manager who grades on a different curve.
- Adjust with reason, not pressure: change a rating only when the evidence or the calibration warrants it, and record why.
Turn it into development
- Convert the review into a concrete plan: what to build, how, and by when — one or two real next steps, not a wish list.
- For anyone at risk, loop in the People Partner; for ratings that drive pay, hand the finalized result to the Compensation Analyst.
Handle disputes
- If the rating is contested, reopen the packet and check it against the rubric and the record — correct it or explain it from the evidence, and escalate a genuine deadlock to the Head of People.
Quality bar
- Every rating traces to observed evidence and a rubric expectation.
- Ratings are calibrated across managers before they're final.
- The review ends in a specific development step, not just a score.
Output format
Review packet: level and period → rubric and goals → per-competency rating with evidence → calibration notes and outliers → development plan (steps, owner, dates) → pay and risk handoffs.
