A performance review is a structured, documented conversation assessing an employee’s contribution over a defined period — usually six or twelve months. Run well, it summarises conversations that already happened, contains no surprises, and ends with agreed priorities for the period ahead.
The performance review has taken a reputational beating over the past decade, much of it deserved. But the periodic review does something the ongoing conversation cannot: it creates a moment to step back and assess a whole period, it produces a documented record, and it forces a decision about what comes next.
The problem was never the review. It was the review being the only performance conversation of the year.
What is a performance review?
A performance review is a structured, documented conversation assessing an employee’s contribution over a defined period. It typically covers achievements against goals, capability demonstrated, development priorities and expectations for the period ahead.
Most follow a similar shape. The employee completes a self-assessment. The manager prepares their own view. The two meet to discuss. The outcome is documented with agreed priorities. Some carry a rating and feed into remuneration; others are purely developmental.
See our glossary entry on performance reviews for the short definition. This guide covers how to run one well.
How often should reviews happen?
Every six or twelve months for most organisations, supported by monthly one-to-ones in between.
| Cadence | Suits | Watch out for |
|---|---|---|
| Annual | Stable workforces, long tenure, low turnover | Reaches only a fraction of a high-turnover workforce |
| Biannual | Most mid-sized organisations | Administrative load if templates are long |
| Quarterly | Fast-changing roles, rapid growth | Review fatigue if it duplicates the one-to-one |
| Anniversary-based | High-turnover frontline teams | Harder to calibrate across a team at one time |
For frontline organisations, anniversary-based reviews often work better than a single annual cycle. A workforce turning over 30% a year has a substantial proportion of people who joined after the last cycle and will leave before the next one.
Preparing for the review
Preparation determines quality more than the meeting itself does. Both people should arrive having reviewed the goals set at the start of the period, the notes from one-to-ones held during it, and any feedback given along the way.
This is only possible if those things were recorded. A review preceded by twelve undocumented conversations is a memory test, and memory reliably over-weights the last six weeks.
Manager preparation
- Re-read goals set at the start of the period
- Review one-to-one notes across the whole period
- Gather specific examples — at least three positive, and any concerns already raised
- Check whether anything in your assessment would be a surprise. If so, that is a problem with the last six months, not the review
- Prepare what you want to say about the period ahead
Employee preparation
- Complete the self-assessment honestly rather than modestly
- Note specific contributions with context, not just outcomes
- Identify what got in the way that the manager may not know about
- Come with a view on what you want to develop next
- Prepare feedback for your manager
Running the conversation
Structure
A workable shape for a 45–60 minute review: start with the employee’s self-assessment rather than yours, discuss goals and what happened to each, cover capability and development, agree priorities for the next period, and close with feedback in both directions.
Starting with the self-assessment matters more than it sounds. It changes the conversation from a verdict being delivered to a comparison of two perspectives, which is both more accurate and considerably less defensive.
Keep the template short
A five to eight question form completed thoughtfully by everyone is worth more than a twenty-question form completed grudgingly by half the organisation. Long templates feel thorough at design time and get abandoned at completion time.
Watch for the two biases
Recency bias weights the final weeks disproportionately. Documented notes across the whole period are the only real antidote.
Halo effect allows a strong impression in one area to colour the assessment of everything else. The check is simple: for each area of the review, can you name a specific example? If not, you may be generalising from a different area.
Letting the review be the first time an employee hears about a performance concern. If something in the review is a surprise, the failure is in the preceding six months of avoided conversations, not in the employee. Our performance conversation guide covers how to raise concerns early enough that they stay conversations.
Should reviews include a rating?
Only if the rating serves a decision. Ratings applied without a purpose create anxiety without adding information.
If ratings feed remuneration or promotion, say so explicitly. Employees should always know which kind of review they are in. Where performance is genuinely below standard, a performance improvement plan is a separate process with different rules. A review that quietly determines pay while presenting itself as a development conversation damages trust in a way that is difficult to repair.
Where ratings are used, calibration matters. Without it, a generous manager’s three is a strict manager’s four, and the data becomes meaningless across teams.
| Approach | How it works | Best for |
|---|---|---|
| No rating | Narrative assessment only | Development-focused organisations |
| Rating without calibration | Manager assigns, no cross-check | Small teams with one manager |
| Calibration session | Managers compare ratings before finalising | Most mid-sized organisations |
| Forced distribution | Fixed percentage in each band | Rarely advisable — damages trust for limited benefit |
Reviews for frontline and casual teams
Standard review processes assume a full-time employee with twelve months of continuous work and a manager who has observed most of it. Frontline reality is frequently different.
Casual and irregular staff. A casual working one shift a fortnight has had perhaps twenty-six shifts in a year, possibly across several sites under different supervisors. A full annual review template is disproportionate. A shorter conversation covering the same ground — how is it going, what do you need, what next — is more honest and more useful.
Multi-site employees. Where someone works across venues, the review needs input from more than one supervisor. Without it you are assessing a third of their work and presenting it as the whole.
Practical constraints. A 60-minute review is difficult to schedule in an operation with no quiet space and no gap in the roster. Thirty minutes, properly prepared, is better than sixty minutes repeatedly postponed.
For high-turnover teams, run reviews on the employee’s anniversary rather than on a fixed organisational cycle. It spreads the administrative load across the year, and it means everyone gets a review at a comparable point in their tenure rather than whenever the cycle happens to land.
After the review
Most of the value in a review is realised afterwards, and most organisations lose it there.
Agreed priorities should appear in the next one-to-one, not resurface at the following review. Development commitments should have a first step with a date. Anything the manager committed to should be tracked as visibly as anything the employee committed to.
The simplest test of whether your review process is working: open a review from six months ago at random. Can you see what happened to what was agreed? If not, the review was an event rather than a decision.
Common mistakes
Surprises. Covered above, and the most damaging failure available.
Reconstructing from memory. If you are writing the review from recall, it will describe the last six weeks and present them as a year.
Rating without evidence. A rating you cannot substantiate with a specific example invites a disagreement you cannot win.
Mixing pay and development without saying so. Employees adjust what they disclose based on what is at stake. Not telling them what is at stake produces a conversation neither person is having honestly.
Filing it and forgetting it. A review that is never referenced again until the next one had no purpose beyond compliance.
Reviews stop being a memory test, because six months of goals, one-to-one notes and feedback are already in one place before anyone opens the form.
Prosper links review cycles to the conversations that preceded them, so managers prepare from a documented record rather than recall. Templates are configurable by role or site, self-assessment and manager assessment run in the same place, and completion is visible across every location.
See how Prosper handles reviewsFrequently asked questions
A structured, documented conversation assessing an employee's contribution over a defined period, covering achievements, capability, development priorities and expectations ahead.
Every six or twelve months for most organisations, supported by monthly one-to-ones. High-turnover frontline teams often suit anniversary-based reviews rather than a fixed annual cycle.
Progress against goals, specific examples of contribution, capability and development priorities, feedback in both directions, and clear expectations for the next period.
Only if the rating serves a decision such as remuneration or promotion, and only with calibration across managers. Ratings without a purpose create anxiety without adding information.
Keep documented notes from one-to-ones across the whole period and prepare from those rather than from memory. Recall reliably over-weights the final weeks.
Use a shorter conversation covering the same ground, and gather input from every supervisor they worked under. A full annual template is disproportionate for someone who worked twenty-six shifts.
The process of comparing ratings across managers before they are finalised, so that a generous manager's rating means the same as a strict manager's.
Agreed priorities should appear in the next one-to-one, development commitments should have a first step with a date, and manager commitments should be tracked as visibly as employee ones.