Understanding Performance Evaluations: Methods, Common Errors and How to Run Them Well
Almost every organization runs performance evaluations, and surprisingly few employees or managers think they work. This guide explains what an evaluation is meant to do, compares the main methods, shows the rating errors that quietly distort results, and gives a practical process and sample wording for managers, plus advice for employees preparing for their own review.
What a performance evaluation is for
A performance evaluation, also called a performance appraisal or performance review, is a periodic process in which an employee’s job performance is documented and assessed against agreed criteria and organizational goals, usually by the immediate manager. That definition hides a tension: the same document is often asked to do three different jobs.
- Development: help the person improve and grow.
- Reward: inform pay, bonus and promotion decisions.
- Record: create documentation that supports decisions, including disciplinary or termination decisions, if they are ever challenged.
These goals pull in different directions. People are rarely open about weaknesses when a rating also sets their bonus, and managers soften feedback when the paperwork will be used against someone. The most useful first question for any organization is which of the three goals matters most, and whether some of them should be handled in separate conversations.
Why so many evaluations disappoint
The evidence on employee reactions is not encouraging. In a 2017 Gallup analysis, only 14% of employees strongly agreed that their performance reviews inspire them to improve, 29% strongly agreed that their reviews are fair, and 26% strongly agreed that they are accurate. About a quarter said their performance was evaluated less than once a year.
There is a deeper problem too. A widely cited review of feedback research by Kluger and DeNisi (1996), covering 607 effect sizes, found that feedback improved performance on average (d = .41) but that over a third of the feedback interventions actually decreased performance. Their explanation is that feedback works less well as attention moves from the task toward the self. Comments about the person (“you are not a team player”) tend to backfire; comments about the work (“the handover notes missed three steps”) are far more likely to help.
Critics also note that annual reviews give feedback too rarely to be useful, rely on subjective criteria, and can breed distrust. One influential idea, the “idiosyncratic rater effect” proposed by Marcus Buckingham and Ashley Goodall, argues that ratings reveal more about the rater than about the person being rated. Whatever its exact size, the point stands: a rating is a human judgment, and the system around it has to control for that.
Some employers have responded by moving away from once-a-year ratings. One estimate by Cappelli and Tavis (2016) was that about a third of U.S. private companies had switched to more informal, frequent manager-employee check-ins. The formal evaluation has not disappeared, but it is increasingly one part of an ongoing conversation rather than the whole of it.
The main evaluation methods compared
| Method | How it works | Strength | Weakness |
|---|---|---|---|
| Graphic rating scale | The manager rates traits or competencies on a 5- or 7-point scale. | Simple and quick; the most commonly used format. | Vague anchors invite leniency and inconsistency between raters. |
| Behaviorally anchored rating scale (BARS) | Each rating point is tied to a described example of job behavior. | Clearer standards; job-specific. | Time-consuming to build for every role. |
| Critical incident log | The manager records specific good and poor behaviors through the year. | Based on evidence, and it supplies material for feedback. | Requires discipline and can feel like surveillance. |
| 360-degree feedback | Input from peers, supervisors and sometimes customers, plus self-assessment. | Multiple viewpoints can reduce leniency and central-tendency bias. | Needs anonymity and training; results can be overwhelming. |
| Ranking and forced distribution | Employees are ranked or sorted into fixed categories on a curve. | Forces differentiation. | Can punish strong teams, damage collaboration and feel arbitrary. |
| Management by objectives (MBO) | Manager and employee agree measurable goals in advance and review results. | Clear targets and employee involvement. | Hard to apply to roles without measurable outputs; may ignore how results were achieved. |
| Self-assessment | The employee evaluates their own performance first. | Surfaces the employee’s view and context the manager may lack. | Self-ratings tend to run high; best used as input, not as the rating. |
No single method is best. Many organizations combine a simple scale for consistency, a few behavioral examples for clarity, and goal-based discussion for direction. The choice should follow the purpose you identified above.
Rating errors and how to reduce them
Because most evaluations rest on managers’ judgment, predictable errors creep in. Knowing them is the first step to controlling them.
| Error | What happens | How to reduce it |
|---|---|---|
| Leniency | Everyone is rated higher than deserved to avoid conflict. | Use behavioral anchors and calibrate ratings across managers. |
| Central tendency | Everyone is rated “average” to avoid difficult conversations. | Require written examples for every rating. |
| Halo effect | One strong trait colors the rating of everything else. | Rate one dimension at a time; use multiple sources. |
| Anchoring | An early impression or last year’s rating drives this year’s score. | Review evidence from the whole period before looking at past ratings. |
| Recency | The last few weeks outweigh the rest of the year. | Keep a running log of achievements and issues. |
| Similarity bias | People who resemble the rater, or are simply liked, score higher. | Focus on observable results and run calibration sessions with other managers. |
One organization-level fix is calibration: managers meet to compare ratings and the evidence behind them, so that a “4” means roughly the same thing in every team. Using more than one source of input, such as peers, customers and hard metrics, also reduces the pull of any single rater’s habits.
What makes an evaluation effective
Gallup’s research suggests three qualities. Effective performance reviews are:
- Achievement-oriented. They acknowledge successes and describe what excellence looks like, rather than dwelling only on faults.
- Fair and accurate. They are held at least every six months, involve the employee in setting goals, and draw on multiple sources of information.
- Developmental. They discuss growth opportunities, such as projects, training or mentoring, alongside performance numbers.
The same Gallup analysis found that only about 3 in 10 employees strongly agree that their manager involves them in goal setting, and that those who do are four times more likely to be engaged. Gallup also stresses that the formal review should be part of an ongoing dialogue, never a replacement for everyday feedback and coaching. That fits with what we know about job satisfaction: people respond to feeling that their work is noticed and that they have a say in how it is judged.
A step-by-step process for managers
Before the review period
- Agree goals and standards early. Write down two to five priorities and what “good” looks like for each. Involve the employee.
- Choose the evidence you will use. Metrics, examples, peer or customer input.
During the period
- Keep a running log of specific achievements, issues and feedback you gave, with dates. This is your defense against recency and anchoring.
- Hold short, regular check-ins. Monthly is a reasonable starting point. Nothing in the formal review should be a surprise.
At the review
- Ask for the employee’s self-assessment first and read it before you finalize your own view.
- Rate against the standards, not against colleagues, unless your system explicitly requires ranking.
- Start with achievements, then discuss gaps using specific examples and their impact.
- Make it a two-way conversation. Leave at least half the time for the employee to respond.
- Agree next steps: goals for the coming period, support needed and a development step.
After the review
- Document the discussion and share a written summary.
- Follow up at the next check-in. Reviews that are never revisited teach people that the process is only paperwork.
Pay is its own topic. If you are deciding how ratings should feed into raises, see developing competitive pay practices, and consider holding the pay conversation separately from the development conversation.
Writing better review comments
Because feedback about the task works better than feedback about the person, the quality of your wording matters. Compare these examples:
| Weak comment | Stronger comment |
|---|---|
| “Needs to be a better team player.” | “In the March release, handover notes were missing for three tasks, which delayed testing by two days. Please complete the handover checklist before leaving a task.” |
| “Great attitude.” | “You stayed calm and kept the client informed when the deadline moved, and the client renewed. Keep doing this.” |
| “Communication needs improvement.” | “Updates in team meetings were often more than five minutes and left out the decision needed. Try leading with the decision and keeping to two minutes.” |
| “Meets expectations.” | “Met all three quarterly targets and exceeded the customer-satisfaction target. The area to grow is delegating routine reporting.” |
A good comment names a specific situation, describes the observable behavior and its effect, and says what to do next. The examples above are illustrative; replace the details with real events from your log.
Fairness and legal basics
Evaluations can create legal risk. Inconsistent or poorly documented appraisals may be used as evidence in discrimination or wrongful-termination claims. A few habits reduce the risk:
- Use the same criteria for people in the same role.
- Base ratings on job-related performance, supported by documented examples, not on personal traits.
- Be consistent between what you write in reviews and what you do in discipline and promotion decisions.
- Train managers so that standards are applied the same way in every team.
This is general information, not legal advice. Employment law differs by country and, in some countries, by state, so check requirements with qualified counsel. The broader compliance picture is covered in our guide to human resource management challenges.
If you are the one being reviewed
- Keep your own record of achievements, with numbers and dates, all year. Do not rely on memory in the weeks before the review.
- Write a self-assessment that matches the criteria your manager will use, and include evidence.
- Prepare for criticism by asking for specific examples and what improvement would look like.
- Bring a development request: a skill, project or training you want, and why it helps the team.
- Follow up in writing to confirm what was agreed.
Giving and receiving feedback well is a skill that grows with practice; see our guide to the essential skills for HR professionals for the wider toolkit.
Frequently asked questions
What is a performance evaluation?
It is a periodic process in which a manager documents and assesses an employee’s performance against agreed criteria and organizational goals. The results are used for development, pay and promotion decisions, and as a record.
What are the most common performance evaluation methods?
Graphic rating scales are the most commonly used, along with behaviorally anchored rating scales, 360-degree feedback, critical incident logs, management by objectives, ranking systems and self-assessment. Many organizations combine several.
How often should performance evaluations happen?
Gallup recommends formal reviews at least every six months, supported by regular informal check-ins. A once-a-year review alone gives feedback too late to change behavior.
What is the halo effect in performance reviews?
It is a rating error in which one strong trait or achievement influences the rating of unrelated areas. Rating one dimension at a time and using evidence from several sources reduces it.
Do performance reviews actually improve performance?
Sometimes. Feedback improves performance on average, but a large meta-analysis found that more than a third of feedback interventions decreased it, especially when comments focused on the person rather than the task. Well-run reviews that are specific, fair and developmental are more likely to help.
Can an employee dispute a performance evaluation?
Most organizations have a way to respond or appeal, such as written comments on the review or a discussion with HR. Check your company’s policy, and keep your own records of achievements to support your case.
Compare HR and management programs.