CSAT versus QA Score: Why Both Move, and Not Together

CSAT asks the customer how the interaction felt. A QA score asks whether the agent did what the scorecard required. They correlate, loosely, and then they diverge on the calls that matter.

When CSAT is high and the call still failed

A friendly agent who skips a mandatory disclosure, offers a discount they cannot honour, or talks a customer into a product that does not fit will often get a good survey. The customer does not know the rule that was broken. If you manage only on CSAT, you will coach people to be likeable and you will miss the breach. This is why a compliance score has to be able to fail on its own.

Collections, insurance and healthcare are full of this pattern. The customer wanted the interaction to end. The regulator wants the sentence that was not said. Industry pages for collections and insurance list the lines that surveys do not see.

When the QA score is high and the customer is not

The form rewarded the script. The customer had to repeat themselves, got transferred, and rated the call poorly. The agent “passed”. Your opening and process criteria are measuring effort that first-contact resolution and the survey are already punishing. Fix the form. Do not average the two numbers into one KPI and call it balanced.

How to use them together

• QA score and compliance score: did we do the job, and did we do it legally.
• CSAT, and the comment if you have one: did the customer feel it.
• A monthly look at calls where the two disagree. Those calls are your calibration material.

Do not pay a bonus on QA score alone. Agents will farm the form. Do not pay a bonus on CSAT alone. Agents will farm the survey and skip the uncomfortable sentence.

Tools that start from the survey, such as Medallia Agent Connect or my.SQM, are built for the customer’s rating. A scorecard system is built for the rule. Mature programmes run both and refuse to blend them.