Short answer: CSAT and NPS tell you how a small, skewed slice of callers felt. They're worth keeping, but they shouldn't be the main way you judge agents. Pair them with sentiment and QA data from far more of your calls.

What's the difference between CSAT and NPS?

CSAT (customer satisfaction) asks how satisfied a caller was with one interaction. NPS (Net Promoter Score) asks how likely they are to recommend you. In a call center, both usually come from a post-call or email survey.

Tom Laird, who runs Expivia and hosts Advice from a Call Center Geek, doesn't pick a side. Back in 2022 he put it bluntly:

"The number one topic, right, is CSAT and NPS. And I'm going to say right now that they're way overrated."

Why CSAT and NPS mislead call centers

Only the extremes answer

"It's only going to be the really happy people or the really ticked off people that are going to really do these surveys anyway."

The middle of your callers, the people you most need to hear from, mostly stay silent.

Agents get punished for following policy

"I've seen reps do an amazing job, but I've lost CSAT scores or NPS scores just because they were following policy on maybe our retention line."

A retention or collections line will always score low on a survey, even when the agent did everything right. That isn't a coaching signal.

What to watch instead (or alongside)

Tom's answer was sentiment scoring. In the same 2022 episode he said his team had "seen an absolute direct correlation between sentiment scores and NPS," and that Expivia keeps sentiment data even for clients who still run surveys.

The point: surveys hear from a few callers, while sentiment and AI-assisted QA can cover far more of your calls. That gives agents feedback they can actually use, and they want it. As Tom put it on another episode:

"I will tell you that the reps ask every single day. Hey, where's my QA?"

For more on this, read why the QA score is dead, how many calls QA should score, and how AI-powered auto QA works. For coaching, see turning QA insights into coaching.

Frequently asked questions

Is CSAT or NPS better for a call center?

Neither is enough on its own. CSAT is closer to a single call, NPS to the overall relationship. Both rely on the few callers who answer surveys.

Why are survey response rates a problem?

The people who respond tend to be very happy or very upset, so the scores swing on extremes rather than typical calls.

Should I stop running CSAT and NPS surveys?

No. Keep them, but don't use them alone to rate agents. Add sentiment and QA data from a much larger share of calls.

Can sentiment analysis replace surveys?

It can cover calls where nobody answers a survey. Tom has said he's seen sentiment scores track closely with NPS, but treat sentiment as one input next to your QA form.

About the author

Tom Laird is the founder of OttoQA and CEO of Expivia, with about 30 years in contact centers. He hosts the Advice from a Call Center Geek podcast. Connect with Tom on LinkedIn.