Measuring Staff Performance: What Each Metric Actually Says
A guide to staff performance measurement: volume and outcome metrics, double attribution, fair comparison and when measurement backfires.
“Who performed well this month?” looks like a simple question. Any system that answers it with a single number produces a team that optimises that number — and it is usually not the thing you wanted.
Three families of metric
Volume
How many calls, tickets, visits, quotes. Easy to measure and easy to game. On its own it shows only busyness.
Outcome
How many sales, resolved tickets, attended appointments. This family is what matters, but it arrives late and is exposed to outside factors.
Quality
Satisfaction scores, first contact resolution, reopened ticket rate. The hardest to measure and the most decisive over time.
A healthy assessment includes one metric from each. A system watching only volume rewards fast, careless work; one watching only outcomes ignores effort and rewards luck.
The double-attribution problem
If two people touched a sale, who gets credit? The person who created the lead, or the one who closed it? No report is reliable until that is answered. There are three approaches:
- Only the closer: simple but unfair. The person generating leads becomes invisible and eventually stops generating them.
- Only the creator: ignores the work of closing.
- Both: the report answers “who contributed what” correctly, but the totals double. That is a deliberate choice and everyone reading the report must know it — otherwise sales totals appear twice the reality.
Whichever you choose, write it down and show it in the report heading. The most expensive mistake is different reports using different methods with nobody aware of it.
Fair comparison
Putting two people’s numbers side by side only means something if their conditions are similar. Common distorting factors:
- Portfolio differences: someone handling enterprise accounts does not work at the same volume as someone handling individuals.
- Territory: in the field, distance directly limits the number of visits.
- Tenure: someone in their third month should not sit in the same table as someone in their third year.
- Quality of incoming work: a rep given poorly qualified leads shows a low conversion rate — and that is not their problem.
The practical answer: compare people against their own history rather than in absolute terms. “Where are they versus last quarter?” is both fairer and more useful than “where are they versus Ahmet?”.
Where measurement backfires
- Handle time targets: push agents to cut conversations short and increase repeat calls.
- Tickets closed: rewards closing early; reopen rates rise.
- Bonuses on satisfaction scores: produce “please give us a 10” and break the metric.
- Visit counts: increase short, purposeless visits.
The common rule: the moment you tie a metric to a target, it stops being a measurement. That is why targets should always be set in pairs — a speed target balanced by a quality target.
The review conversation
Numbers are the start of a conversation, not its conclusion. A structure that works: let people see their own numbers beforehand and interpret them first in the meeting. A review where the manager reads out figures produces defensiveness; one where the person interprets their own data usually surfaces the real cause.
Frequently asked questions
How many metrics should we track?
Three to five per person. More scatters focus, and the team cannot tell which number matters.
Should performance data be shared with the team?
Everyone’s own data, always. Team rankings depend on culture: motivating in some teams, corrosive to collaboration in others. Try it with a small group before deciding.
How should a new joiner be assessed?
On learning and process adherence rather than outcomes. In the first three months measure “keeping accurate records” and “following the process”; expecting results is premature.
Next step: to see how staff evaluation reports come out, request a demo.