Performance scores and the talent grid
The talent grid places people by how they are performing and which way they are moving. It turns a list of ratings into a view you can act on — provided you know what the bands actually mean.
Where: Performance › Reports You need: performance rights covering the employees shown
How a rating becomes a score
Ratings run 1 to 5. The grid converts them to a percentage by multiplying by 20:
| Rating | Score |
|---|---|
| 1 | 20 |
| 2 | 40 |
| 3 | 60 |
| 4 | 80 |
| 5 | 100 |
The three performance bands
| Band | Score | Rating |
|---|---|---|
| Below | under 60 | below 3 |
| Meets | 60 to 79 | 3 to just under 4 |
| Exceeds | 80 and above | 4 and above |
💡 The boundaries land exactly on whole ratings. A 3 is the bottom of Meets, and a 4 is the bottom of Exceeds. There is no partial credit between them unless your ratings carry decimals — so a reviewer choosing between 3 and 4 is choosing between two bands, not nudging a number.
Tell reviewers this. Most rating inflation comes from people treating 4 as "good" without realising it moves someone into Exceeds.
Trend
Trend compares an employee's score at the start of the window against the end:
| Trend | Change |
|---|---|
| Improving | up 3 points or more |
| Declining | down 3 points or more |
| Steady | within 3 points either way |
Three points is well under one rating step, so trend picks up genuine movement without treating ordinary variation as a signal.
Reading the grid honestly
The grid is only as good as the ratings behind it, and the ratings come from assessments and goals. Before drawing conclusions from it:
- Check your assessments are actually discriminating — multi-select questions award full marks to everyone and flatten the distribution. See How assessment scores are calculated.
- Check goal priorities are used. If every goal is Critical, weighting is doing nothing.
- Check reviewers rated consistently. One generous reviewer moves their whole team a band.
What Happens Next
- The grid reflects ratings in the period you select, with trend measured over its own window.
- It is a report, not a record — it changes as ratings change.
- Nothing in KAMI acts on a grid position. What follows is a management decision.
Tips
- Brief reviewers on the band boundaries before a cycle, not after. Knowing that 4 means Exceeds changes how carefully people rate.
- Look at the distribution before the individuals. If nearly everyone is in one band, the problem is the assessment design, not your workforce.
- Treat trend cautiously over short windows. Three points is a small movement, and one cycle is not a pattern.
- Never use the grid as the sole input to a pay or promotion decision. It summarises ratings; it does not know context.
- Check who is missing. Anyone without ratings does not appear, and absence from the grid is not a performance signal.
Troubleshooting / FAQ
Q: Everyone is in Meets. Ratings are clustered at 3. Check whether reviewers are defaulting to the middle and whether the assessment discriminates at all.
Q: An employee rated 4 shows as Exceeds — is that right? Yes. 80 is the bottom of Exceeds, and 4 × 20 = 80.
Q: Someone is missing from the grid. They have no ratings in the period.
Q: The trend says steady but the score clearly moved. It moved by less than 3 points.
Q: Two teams look very different. Is one better? Possibly, or their reviewers rate differently. Compare reviewer behaviour before concluding anything about the teams.
Screenshots
These screenshots came from our previous help centre and may show an earlier version of the interface.
