Churn is the only honest satisfaction metric. Why I stopped trusting NPS.

You already have a better number

The single biggest mistake with CSAT and NPS is using the score as a proxy for how happy your customer base is — because you already have a truer one. Logo churn and revenue churn are the actual truth of whether customers are satisfied enough to stay and pay. A customer can hand you a 9 on an NPS survey and leave three months later; the churn number doesn't lie the way a survey does. Whether customers stay and keep paying is satisfaction in the only form that matters to the business. The score is a weak, biased, lagging stand-in for it.

NPS was easy to collect and hard to act on

That's the honest verdict on company-wide NPS. It rarely told us anything we could actually do something about. Occasionally an individual detractor response let us catch one customer having a bad experience and intervene — that's real value, and worth having. But as an aggregate number on a board deck, it mostly just sat there. A single figure that summarizes thousands of relationships doesn't point you at any specific action, and a metric you can't act on is a metric you're collecting for the comfort of having it.

CSAT's real value is coaching agents

CSAT I found more useful, but not for what people expect. It wasn't a great read on overall customer happiness — but it was genuinely useful for one thing: coaching support agents. A ticket could be resolved in great time and still score low because the agent's tone or attitude was off. That gap — fast resolution, low score — is exactly the coaching signal a new agent needs. So CSAT earns its keep at the agent level, as a training tool, not as a verdict on the business. Use it to develop your team; don't mistake it for a satisfaction gauge.

The comment beats the score

If you're going to run these surveys at all, the value isn't in the number — it's in the free-text comment. The score is blunt; the comment is where customers actually tell you something. We aggregated the comment data to identify trends — the recurring asks, the repeated complaints, the themes a single digit can never show. Treat the score as the thing that got someone to write the comment, and put your analysis into the words. A team that watches the NPS trend but never reads the comments is throwing away the only genuinely actionable output the survey produces.

What's actually better than a survey? Talking to customers.

A survey is a thin, biased substitute for a conversation — only the delighted and the furious answer, the quiet middle stays silent, and the whole thing is gameable on top of that. The richest understanding of your customers comes from the real thing: talking to them, regularly, at every level. Sales, success, and support hear things every day. But most importantly, senior executives should be doing real customer visits.

The companies that truly understand their customers aren't the ones with the highest NPS — they're the ones whose leadership is in front of customers often enough that the score is just a footnote to what they already know. Build that rhythm and the survey becomes a sanity check, not your primary source of truth. Read the score as a soft pulse, read the comments for signal, and read your churn rate as the verdict. Then go talk to the people behind all three.