What Is a Good Call Centre QA Score?

People ask for a benchmark because a benchmark feels like an answer. “Good is 90” is not an answer until you know what the form measures and whether two analysts would both give a 90.

Why copied targets mislead

Scorecards are not standardised. A form with three soft-skills lines and no fatal criteria will sit higher than a form that zeros the call for a missed disclosure. Comparing your 86 with another company’s 92 is comparing two different tests. Vendors who publish “our customers average 94” are telling you about their customers’ forms.

A target you can defend

1. Run calibration until analysts agree on the critical lines.
2. Score a baseline month without telling the floor a target. You need the current distribution, not the distribution of people chasing a number.
3. Set the target near the upper part of that distribution, not at 100. A target nobody reaches becomes wallpaper. A target everyone reaches without a behaviour change was already true.
4. Set a separate, stricter line for compliance: the fatal rate you will not tolerate. Often that number is zero on specific criteria, which is a different statement from “QA score above 85”.

What to do with someone below the line

Look at the criteria, not the integer. An agent at 78 because openings are sloppy is a coaching plan. An agent at 78 because of one fatal disclosure is a compliance conversation. The same integer should not produce the same meeting.

If you need a sentence for a steering committee: a good QA score is a stable score, on a calibrated form, with compliance reported beside it. The integer is local. The method is not.

The KPI pack this sits in is the QA KPI guide.