Employee performance evaluation is the process of measuring results and behaviors against criteria agreed up front, not against accumulated impressions. It is not a report card; it is a tool for mutual growth.
Research on procedural justice indicates that a procedure is perceived as fairer when it is applied consistently across people and over time and when decisions rest on accurate information; in a classic study of performance evaluations judged fair or unfair by the managers who had received them, consistent application of standards is among the determining factors [1]. The finding points to an opportunity more than a judgment.
The next sections cover criteria, indicators, the structure of the review meeting and common mistakes.
Define evaluation criteria before looking at results
A useful evaluation starts from criteria agreed up front, not from results observed after the fact. Deciding afterward what "really counted" is the most common cause of evaluations perceived as unfair. Transparent criteria — few, explicit, shared — reduce the perception of arbitrariness well before the numbers do.
How many evaluation criteria can a company handle without turning them into a complicated grid? Adding criteria at evaluation time tells the team member that "the rules were different."
A practical, working grid includes no more than 5 criteria: 3 for results and 2 for behaviors.
Results criteria (3). They concern what the team member produced: goals achieved, outputs delivered, quality of work against the agreed standards. They must be specific and measurable — not "did a good job" but "closed 8 deals out of the 10 planned" or "delivered reports on time in 90% of cases."
Behavioral criteria (2). They concern the way the team member worked: collaboration with the team, communication, problem handling, adherence to operating practices. They are no less important than results — they are often more predictive of future performance. Nor are they a niche detail: in the 2024 Excelsior survey, flexibility and teamwork were rated as highly important for 67% and 58%, respectively, of the hires planned by Italian companies, ahead of the ability to work independently and problem solving (42-43%) [3].
Criteria should be shared with the team member at the start of the period, not at evaluation time. A team member who does not know the criteria they will be evaluated on is working in the dark, and any evaluation will seem arbitrary no matter how sound it is.
Linking to the job description is useful to anchor behavioral criteria to the responsibilities of the role: if a job description exists, the evaluation criteria are a natural extension of it (see also how to build a job description).
Distinguish results from behaviors without confusing them
A good team member can achieve results in ways you would not want repeated; another can work with impeccable method and still miss the target. Confusing the two levels — results and behaviors — produces evaluations that do not guide growth. Keeping them separate lets you recognize and correct what really matters.
What should you do when results are good but behaviors are not, and vice versa? Evaluating only the numbers is the fastest way to entrench the behaviors you do not want.
The 2×2 results/behaviors matrix defines four quadrants, each with a corresponding action.
| Appropriate behaviors | Problematic behaviors | |
|---|---|---|
| Results achieved | Recognition and development | A conversation about method, not results |
| Results missed | Analysis of obstacles, support plan | In-depth assessment of continuing the working relationship |
The hardest quadrant to manage is the top right: good results but problematic behaviors. Ignoring the problem so as "not to disturb a high performer" is a mistake you pay for in the medium term, when the behavior becomes entrenched or spreads as a model. The conversation in that case is not about the result — which should be recognized — but about the method used to achieve it.
The bottom-left quadrant (appropriate behaviors, missed results) instead calls for a diagnosis of the obstacles: often the problem is not the person, but the conditions they work in, the information they received or the goals they were assigned.
Choose indicators the team member can actually influence
Measuring a team member on variables that depend on external factors or on other people's choices breeds distrust, not direction. It is also the point at which evaluation stops guiding anything: an indicator the team member cannot move tells them nothing about what to do differently next month. The question is not "what to measure," but "what is reasonable to ask."
Which indicators is it fair to evaluate a salesperson, a technician or a back-office employee on? Measuring a team member on a figure they do not control is an evaluation that undermines itself.
The controllability principle applies differently by role.
Sales. The controllable indicators are: number of contacts made, quotes prepared, appointments set. Final revenue also depends on external variables (the market, customer decisions, pricing policy): it can be a goal but not the only criterion.
Technical. The controllable indicators are: compliance with technical specifications, completeness of documentation, execution times against the plan. Delays caused by third parties or by changes in specifications cannot be attributed to the technician.
Back office/administration. The controllable indicators are: accuracy of the data produced, meeting internal deadlines, handling exceptions. Quality often also depends on the information received upstream: here too the controllability principle must be applied precisely.
To build a system of indicators consistent with your business strategy, the article on business KPIs offers a useful methodological framework.
Structure the review meeting in 4 clear parts
The review meeting, stripped to the essentials, follows a four-part sequence: a shared reconstruction of the period, a reading against the agreed criteria, listening to the team member's perspective, and setting priorities for the next period. Skipping one of the parts — typically the listening — produces meetings perceived as monologues. The order matters: reconstructing before evaluating lowers defenses.
How long should a useful review meeting last? A thirty-minute meeting with no initial reconstruction is a judgment in disguise.
The four parts, with indicative timing for a 60-90 minute meeting.
1. Shared reconstruction of the period (15-20 minutes). You go back over the main events of the period — successes, difficulties, changes in context — together, not unilaterally. The evaluator describes their reading, then asks for the team member's reading of the same events. This step reduces recency bias (remembering only the last few months) and the halo effect (one positive event coloring the whole period).
2. Reading against the agreed criteria (20-25 minutes). You compare the observed results with the criteria agreed at the start of the period. The reading is descriptive before it is evaluative: "on this criterion the results were X, against an expected Y." Judgment comes after description, not before.
3. Listening to the team member's perspective (15-20 minutes). Open questions about the team member's reading: what worked from their point of view, what got in their way, what they would change. This part is not a concession to organizational populism: it is a source of information that lets you separate performance problems from problems of context.
4. Setting priorities for the next period (10-15 minutes). Agreement on 1-2 concrete priorities for the coming months, with any resources or support needed. Without this step, the meeting ends with no view of the future.
For the technical structure of feedback during the meeting, the article on how to give feedback to your team explores the communication side in depth.
Turn the evaluation into measurable growth goals
An evaluation that does not turn into priorities for the next period ends with the meeting itself. One third of Italian companies with at least 10 employees report that some of their staff lack the skills required by their job [2]: that is exactly the gap a review meeting can name and turn into a goal, instead of leaving it unspoken until the following year. Turning it into a goal takes the discipline of writing down "what changes starting Monday."
Which growth goals really hold up as the output of a review meeting? A review meeting with no follow-up plan is a meeting that ends at the table.
A practical structure for post-evaluation goals: 1-2 growth goals and 1 enabling resource.
Growth goal (1-2). Phrased as "be able to do X in context Y by date Z." Not "improve communication" but "handle the briefing with type A customers independently by the second quarter." Specificity is the minimum condition for the goal to be assessable at the next meeting.
Enabling resource (1). What the organization provides to support the goal: shadowing, access to a tool, a change in responsibility for certain tasks. A goal without an enabling resource is a one-sided burden on the team member.
To frame growth goals within a structured training plan, the article on how to build a corporate training plan offers the methodological framework.
Make evaluation recurring without weighing down the calendar
An isolated annual evaluation piles up tensions that the meeting cannot release. Adding short check-ins — quarterly or every six months — spreads out the dialogue and reduces the "exceptional event" effect. The cadence is not a formality; it is how you design the working relationship.
What evaluation cadence holds up without becoming an extra burden? A formal evaluation every twelve months is a conversation squeezed into two hours a year.
A distributed cadence that works in practice:
Structured annual evaluation. The full meeting in 4 parts, with criteria, reading and goals. It requires preparation from both sides and a written output. It can last 60-90 minutes.
Light quarterly check-in. A short conversation (20-30 minutes) that updates the status of the agreed goals, flags any obstacles and checks whether conditions have changed. It does not require a formal document, but a short shared note is useful.
The cumulative effect of the quarterly check-in is significant: it reduces surprises in the annual meeting, increases the team member's sense of being followed and supported, and allows course corrections along the way instead of after the fact.
To fit review meetings into the recurring moments already in place in your company, the article on how to run effective business meetings offers a complementary perspective.
Recognize and correct common mistakes in performance evaluation
The most common mistakes in evaluation are not about technique but about posture: criteria not shared up front, confusion between results and behaviors, uncontrollable indicators, the monologue meeting, evaluation with no follow-up goals, an annual-only cadence. Recognizing them is more useful than memorizing them, because they show up in different forms depending on the size of the company. Addressing them takes honesty more than method.
Which mistake costs more: evaluating too little or evaluating with the wrong criteria? Both extremes produce the same outcome — team members who do not know what they need to do better.
The six most recurrent mistakes, with a small correction.
1. Criteria defined at evaluation time. It creates a perception of arbitrariness. Correction: share the criteria with the team member at the start of the period, in writing.
2. Confusion between results and behaviors. It leads to rewarding a result achieved the wrong way or penalizing the right method that missed the target. Correction: apply the 2×2 matrix as a reading framework before the meeting.
3. Indicators outside the team member's control. It breeds distrust in the system. Correction: check for each indicator whether the team member can directly influence it.
4. The monologue meeting. The team member has no room for their own reading. Correction: explicitly plan the listening part into the structure of the meeting, with prepared open questions.
5. Evaluation with no follow-up goals. The meeting ends with no view of the future. Correction: close every evaluation with 1-2 specific goals and 1 enabling resource.
6. An annual-only cadence. Tensions build up and the meeting becomes a venting ritual. Correction: introduce light quarterly check-ins as part of the process.
Limits and conditions of applicability
The method described in this article assumes that roles are defined well enough to allow specific criteria to be formulated. In startups or in organizations with very fluid roles, formal evaluation may be premature: in those cases, it is more useful to focus on frequent feedback conversations than on a rigid evaluation structure.
The institutional sources cited (ISTAT, Unioncamere-Excelsior) describe Italian companies as a whole, not only small and mid-sized ones, and the research on procedural justice brings together international studies on organizations of very different sizes. The findings on the link between explicit, consistently applied criteria and perceived fairness [1] should be considered indicative evidence, not a guarantee of effect. The organizational and cultural variables specific to each company significantly influence the effect of evaluation practices.
This article focuses on individual evaluation (a manager evaluating a team member). 360-degree evaluation, peer review and other participatory formats follow their own logic, which is beyond the scope of this text.
FAQ
How far in advance should evaluation criteria be prepared? Ideally, criteria are defined at the start of the period — at the beginning of the year or the evaluation quarter — and shared with the team member in a short alignment conversation. A complex formal document is not necessary: even a short note shared by email is enough to create the reference point.
How do you evaluate a very senior team member who has more experience than the evaluator? In those cases, the evaluation focuses more on the agreed goals and results indicators, and less on behaviors. The listening part becomes especially important: senior team members often have a more accurate reading of their own performance than the evaluator.
Is it necessary to use a numerical scale for evaluation criteria? Not necessarily. Numerical scales create an impression of precision but often lead to arguments over a single point rather than over what matters. In many companies, a structured narrative evaluation ("on this criterion the result exceeded expectations; on this other one there is room for improvement") works better and generates more useful conversations.
What should you do if the team member disputes the evaluation? A dispute is often a sign that the criteria were not clear up front, or that the listening part did not get enough room. Before responding to the dispute, it is useful to check whether the team member is right about at least part of what they raise: a dispute often contains useful information about the quality of the process.
Operational summary
Employee performance evaluation produces concrete results when it rests on criteria agreed up front (not built retroactively), distinguishes results from behaviors with the 2×2 matrix, uses indicators the team member can actually influence, structures the meeting in four sequential parts with room for listening, turns into 1-2 growth goals with enabling resources, and is spread over a structured annual cadence plus light quarterly check-ins. Without even one of these elements, evaluation loses part of its effectiveness as a tool for mutual growth.
Conclusion
Employee performance evaluation is not a report card; it is a tool for mutual growth: a few criteria agreed up front, a distinction between results and behaviors, indicators the team member can actually influence, a meeting structured in four parts, translation into growth goals, and a cadence that spreads out the dialogue instead of concentrating it.
The thread that ties these elements together is the difference between evaluating to judge and evaluating to guide. The first ends with the meeting; the second opens the next period. To frame feedback conversations outside the evaluation moment, also read how to give feedback to your team and, on the growth side, how to build a corporate training plan.
An organization that evaluates with method is one where team members know what is valued and what is not, promotions and changes in responsibility become understandable, and merit stops being perceived as arbitrary. It is a more solid organizational posture and a more sustainable path to internal growth, within reach of organizations of any size.
Sources and references
[1] Bobocel D. R., Gosse L., "Procedural Justice: A Historical Review and Critical Analysis", in Cropanzano R. S., Ambrose M. L. (eds.), The Oxford Handbook of Justice in the Workplace, Oxford University Press, 2015. Available at: https://uwaterloo.ca/fairness-at-work-lab/sites/default/files/uploads/files/procedural_justice_an_historical_review_and_critical_analysis.pdf
[2] ISTAT, "Formazione nelle imprese — Anno 2020", Italian National Institute of Statistics, December 30, 2022. Available at: https://www.istat.it/it/files//2022/12/REPORT-formazione-imprese.pdf
[3] Unioncamere — Italian Ministry of Labor and Social Policies, Excelsior Information System, "La domanda di professioni e di formazione delle imprese italiane nel 2024", Unioncamere, 2025. Available at: https://excelsior.unioncamere.net/sites/default/files/pubblicazioni/2024/Domanda_professioni_formazione_imprese.pdf