Review your team fairly — and with data

How ratings become a percentage

What most costs an appraisal its credibility is not the rating an employee received but nobody being able to explain how it emerged. Someone who asks "why three and not four?" and gets an impressionistic answer leaves convinced the appraisal is a personal opinion written on a form.

The calculation itself is simple and transparent: a rating is picked for each competency from a defined scale, converts to a fraction of the top rating, and is multiplied by the competency's weight to give its earned points. The employee's percentage is total earned divided by total weights times a hundred, then mapped to a final rating per the scale's bands.

The value of that clarity is shifting the conversation from "why is my rating this?" to "why is my rating on this specific competency this?" — the second can be answered with examples and the first cannot.

Weight prevents false equivalence

A five-competency form with no weights says they are equally important, which is almost never true. Treating punctuality as equal to quality of work tells the team the two are equivalent — a message no company wants to send.

How weights are distributed is the clearest statement management makes about what it values — clearer than any memo. A convenient practical detail: the weights need not sum to a hundred, because the percentage divides earned points by whatever their total is. So a competency can be added or removed without redistributing every weight — which would otherwise have made any change impractical.

Why self-assessment is not merged with the manager's

The employee assesses themselves, then their manager does, and the temptation is to average the two. But merging destroys the most important information in the whole exercise: the gap between them.

Someone rating themselves far above their manager does not necessarily have a performance problem — they may have an expectations problem nobody clarified. Someone rating themselves far below may be underselling good work. Both need a conversation, and neither shows in an average.

So both are stored side by side, the gap stays visible and becomes material for discussion instead of dissolving into a single figure that says nothing about where it came from.

Closing in two stages

If both stages were open together, the thing that ruins self-assessment entirely would happen: the employee sees their manager's rating and adjusts their own to match. Self-assessment then becomes an exercise in guessing rather than reflection.

So the ordering is deliberate: the period opens and the roster of who is due is prepared, the employee completes their self-assessment, then it closes and can no longer be edited, then managers begin their review. Once review starts there is no going back — because going back after seeing the rating is exactly what the ordering existed to prevent.

The review is the only approval gate: no employee acknowledgement step and no chain of approvals. The system enforces no pass mark — the pass line is the company's decision, not the software's.

Peer appraisal: more than one view

One person's view of a colleague's performance carries what it cannot avoid: they see part of the work, not all of it. Peer appraisal gathers evaluations from several raters for one employee, and the combined result is computed as an overall percentage across them.

The settings define who rates whom each month, defaulting to the direct manager and the department manager and adjustable. Each rater fills the same competency form — so comparison stays possible and no two raters are measuring different things.

Development plans: written, not derived

The individual development plan and the career-path plan are separate tools recording strengths, development gaps, activities, goals and training needs; each passes an accept-or-refuse decision then locks.

The system does not build them automatically from appraisal scores — a choice rather than a gap. A score says what the level of performance is, not why: someone rated low on a competency may need training, or tools, or clarity about expectations nobody explained. All three produce the same number and call for three different plans.

The result is used as a reference to write against, and the training need links through to Learning & Development — turning the plan from a page of intentions into a scheduled course with a date.