Performance Review

A performance review evaluates how an employee performed over a defined period against defined expectations, records the result, and feeds decisions about pay, development, promotion, and sometimes continued employment.

The review is the documentation event. The performance management that matters happens continuously between reviews, through goals, feedback, and one-on-ones. A review that surprises an employee is a symptom of a management problem rather than a working process.

How a cycle works

A cycle begins with a period and a population: who is being reviewed, over what dates, and who is excluded because they are too new. Most organizations exclude anyone hired within the last sixty to ninety days of the period and review them on a start-date anniversary instead.

Inputs are gathered next. The employee usually writes a self-assessment, peers or cross-functional partners may contribute feedback, and the manager pulls together goal outcomes and their own observations. The manager then drafts an assessment and a proposed rating.

Calibration comes before delivery, not after. Managers in a group review their proposed ratings together against the definitions and against one another, and adjust so that a given rating means roughly the same thing regardless of who the manager is. Only after calibration is the rating final.

Delivery is the conversation: what the rating is, what evidence supports it, and what changes for the next period. Compensation decisions follow, and are usually communicated separately so the development discussion is not swamped by the number.

Cadence varies. Annual cycles with a mid-year check-in remain the most common, quarterly cycles are used where work and priorities move quickly, and many organizations run lightweight continuous check-ins alongside a single formal annual documentation event.

A rating scale that actually differentiates

Scale design determines whether ratings carry information. Five points with written definitions is a common and workable choice.

What calibration is for

Calibration is the session where managers compare proposed ratings before anything is communicated. It is the step most often skipped and the one that does the most work.

• It makes ratings comparable across managers, so a rating from a lenient manager and a demanding one mean approximately the same thing.
• It exposes rating inflation, since a group where nearly everyone exceeds expectations is visible immediately when the group is viewed together.
• It forces evidence. A manager who cannot describe what a rating is based on usually revises it in the room.
• It surfaces pattern differences worth examining, including whether ratings vary by group in ways that performance evidence does not explain.
• It builds a shared standard over time, which is the durable output. The individual adjustments matter less than the calibrated judgment managers carry into the next cycle.
• It should happen before ratings are shared with employees. Adjusting a rating after it has been delivered destroys trust in the process.

Where review cycles break

The mechanics are simple. The failure modes are predictable.

• Recency bias. Managers weight the last six weeks, which is why contemporaneous notes throughout the period matter more than the form.
• Rating inflation. Avoiding hard conversations produces a distribution where almost everyone is above average, which makes pay differentiation impossible to justify.
• Forced distribution applied to small teams. Requiring a fixed percentage of low ratings in a team of six manufactures a low performer who may not exist.
• Ratings that arrive as news. If the review is the first time an issue is raised, the employee reasonably concludes it was invented to justify a decision.
• Goals set once and never revisited, so the employee is evaluated against priorities that changed in month three.
• Evaluating personality rather than performance. Comments about attitude, style, or being a culture fit are weak signal and, when they vary by group, are a liability.
• Pay and development crammed into the same conversation, where the number reliably drowns out everything else.
• Documentation that hedges. Reviews softened to avoid a difficult conversation become the record that contradicts a later termination decision.

Why it matters operationally

The review cycle is where the organization decides what performance means, and that decision propagates into pay, promotion, succession, and separation. If the ratings do not differentiate, every downstream decision has to be made on some other basis, usually manager advocacy, which is less consistent and much harder to defend.

The cycle is also the only regular moment where every employee is guaranteed a documented conversation about their work. That guarantee is worth protecting even in organizations that have moved most feedback to a continuous model, because continuous feedback tends to reach the people whose managers were already good at it.