The work behind the count

Follow a case · fictional example

The dashboard sees eight ahead of four

Two colleagues spend one hour resolving requests. The dashboard uses the number closed to rank their performance.

The same working time: 60 minutes

Complex requests

4

More time is needed to resolve each case.

Routine requests

8

More cases can be closed in the same hour.

The count describes volume. A bonus based on it also creates a reason to avoid difficult cases.
When an accent becomes a performance score

Now suppose the same fictional dashboard also imports customer ratings. One customer gives a worker a low score because of their accent, despite the problem being resolved. Once that rating enters the average, the dashboard presents it as a judgement of performance.

An intersectional reading follows how a judgement associated with language or perceived origin becomes an employment consequence, and whether the worker’s contractual position leaves room to challenge it. The organisational question is how the rating is checked before it affects access to work.

The system can count correctly while giving an incomplete account of performance. The number represents completed requests; it says little about their difficulty or whether a solution will last. A rushed closure followed by another complaint could even appear more productive than a longer conversation that resolves the issue.

Once a measure influences rewards, it can also influence decisions about the work. In this example, the organisation wants difficult cases resolved, while its reward rule makes routine cases more attractive. The problem involves the incentive attached to the count as well as the count itself.

A small numerical difference can have a large consequence

Consider a fictional bonus rule that pays at 80 points and pays nothing at 79. The one-point difference is small on the scale, but it decides the entire bonus. The threshold gives that difference a significance that the score alone does not reveal.

Reviewing an evaluation therefore involves following the number into the decision. Removing an inaccurate entry may change the score. Revising a threshold or recognising time spent on difficult work changes how the organisation uses it.

Changing the rating form can change the outcome

Customer ratings introduce another design choice: what customers are asked to judge and how they can express that judgement. A rating form can give different amounts of room for personal expectations to enter an assessment.

The same service can be judged differently

A rating also reflects an encounter between people. When an organisation uses customer judgements to allocate work, it needs to consider whether expectations about workers affect those judgements. Combining them into an average does not remove that question.

Keeping the purpose of the work in view

Return to the colleague handling the four difficult requests. Adding more indicators will help only if the evaluation can recognise why those cases took longer and why resolving them mattered. The same applies to helping a new colleague or preventing a mistake before it reaches a customer.

An appraisal that overlooks these contributions can repeatedly disadvantage the people who make them. A useful performance system leaves room to explain the work, correct omissions and reconsider incentives that make necessary tasks harder to take on.

Sources