Performance Review System Template

A 4-point scale forces reviewers to make a meaningful distinction rather than defaulting to the middle. Research from [Deloitte's performance management studies](https://www2.deloitte.com/us/en/insights/focus/human-capital-trends.html) shows that 5-point scales cluster heavily at 3 and 4, reducing signal quality. A 4-point scale with clear behavioral anchors for each level produces more actionable differentiation. Track your rating [distribution metrics](/glossary/aarrr-pirate-metrics) to verify that the scale is working as intended.

Build bias detection directly into the calibration tool. Show rating distributions by gender, ethnicity, tenure, and team alongside the calibration grid. Require managers to write a 2-3 sentence justification for any "Exceeds" or "Does Not Meet" rating before the calibration session. Train calibration facilitators to flag patterns where a demographic group is consistently rated lower than peers with similar goal achievement.

For growth-focused peer feedback, anonymity increases candor. For formal review inputs that affect ratings and compensation, transparency builds accountability. A common pattern is anonymous upward feedback (reports evaluating managers) and attributed lateral feedback (peers). Let your company culture and size guide the decision. Under 50 employees, true anonymity is difficult to maintain.

Run calibration sessions via video with a shared calibration dashboard. The facilitator shares the screen showing the 9-box or rating distribution view. Each manager presents their team's ratings with supporting evidence. Async pre-work is essential: require managers to submit written justifications 48 hours before the session so calibration does not devolve into live essay-writing. Use your [product analytics](/analytics-guide) to track calibration session duration and rating change frequency.

Start with three features: a self-review form, a manager review form with side-by-side self-review visibility, and a simple [dashboard](/glossary/aarrr-pirate-metrics) showing completion rates. Skip calibration tooling, 360 feedback, and LMS integration in v1. These can be done manually or in spreadsheets while you validate the core review workflow. Use the feedback from the first review cycle to prioritize v2 features with the [RICE framework](/frameworks/rice-framework). ---