← ccPlanning Academy · Metrics track

What a metric is for

Free visual lesson · about 5 minutes · short quiz at the end

ccPlanning academy · metrics

What a metric is for

Choosing what to measure is choosing how people will behave.

The big idea

You get the behaviour you measure.

A metric isn’t a neutral mirror — it’s an instruction. The moment a number has a target attached, people optimise for the number. So the first question about any metric isn’t “is it accurate?” but “what behaviour will chasing it produce?”

Goodhart’s law

“When a measure becomes a target, it stops being a good measure.”

Target AHT and calls get rushed or transferred. Target calls-per-hour and quality slips. The number improves while the thing it was meant to represent quietly gets worse. Every contact-centre metric is vulnerable to this.

The real purpose

A metric exists to inform a decision.

If a number doesn’t change what someone does, it’s decoration. Before adopting a metric, name the decision it serves: what would we do differently if it were high versus low? No decision, no metric.

Two failure modes

Measuring the wrong thing — or too much.

The wrong metric drives the wrong behaviour. Too many metrics drown the few that matter and let people cherry-pick whichever looks good today. Both are common; both are choices you can avoid.

The discipline

Measure outcomes, watch the drivers.

Hold people to the outcome that matters (served customers, good experience) and use the operational metrics (AHT, occupancy, adherence) as diagnostics — things you watch to understand, not targets you beat. That distinction runs through this whole track.

Goodhart in one shift

Target AHT, watch what happens

Tell agents AHT must drop and it will — calls get rushed, customers cut off mid-sentence, tricky queries quietly transferred to someone else. The number on the dashboard falls. Resolution falls with it.

A week later those rushed contacts come back as repeats, and total workload is up. The measure improved; the thing it stood for got worse. That’s the whole trap.

The takeaway

Pick metrics for the behaviour they create.

Every measured number is an incentive. Choose the few that drive the right behaviour and serve a real decision; treat the rest as diagnostics, not targets. The lessons ahead apply this to the metrics you use every day.

Now test yourself ↓

1 / 7

Slides done? Here’s the same idea in a bit more depth — the part worth keeping.

In depth: choosing what to measure is choosing how people behave

A metric isn’t a neutral mirror — it’s an instruction. The moment a number has a target attached, people optimise for the number, so the first question about any metric isn’t “is it accurate?” but “what behaviour will chasing it produce?” You get the behaviour you measure, which means every dashboard is quietly a statement about what you want people to do.

Goodhart’s law is unavoidable

“When a measure becomes a target, it stops being a good measure.” Target AHT and calls get rushed or transferred; target calls-per-hour and quality slips; the number improves while the thing it was meant to represent quietly gets worse. Every contact-centre metric is vulnerable to this, which is why the purpose of a metric matters so much: a metric exists to inform a decision, and if a number doesn’t change what someone does, it’s decoration. Before adopting one, name the decision it serves — what would we do differently if it were high versus low? No decision, no metric.

Two failure modes, one discipline

There are two common, avoidable mistakes: measuring the wrong thing, which drives the wrong behaviour, and measuring too much, which drowns the few metrics that matter and lets people cherry-pick whichever looks good today. The discipline that avoids both is to measure outcomes and watch the drivers: hold people to the outcome that genuinely matters — served customers, good experience — and use the operational numbers like AHT, occupancy and adherence as diagnostics you watch to understand, not targets you beat. That single distinction runs through this whole track.

The principle to remember: pick metrics for the behaviour they create. Choose the few that drive the right behaviour and serve a real decision; treat the rest as diagnostics, not targets.

Quick quiz

Five questions. Pick an answer to each, then check your score.

1. What’s the first question to ask about a metric?

A metric is an instruction — people optimise for the number, so behaviour comes first.

2. What does Goodhart’s law say?

Target AHT and calls get rushed; the number improves while the real thing gets worse.

3. When is a metric just ‘decoration’?

A metric exists to inform a decision — no decision, no metric.

4. What’s the problem with tracking too many metrics?

Too many metrics dilute focus and enable selective reporting.

5. What’s the recommended discipline?

Outcomes are the target; AHT, occupancy and the like are things you watch to understand.