There's no generally good number, and the circulating benchmarks are hard to use because they rarely state the method and the country. What is usable is the movement of your own number under an unchanged setup, with the response count attached. An NPS going from +12 to +19 across four hundred responses says more than any industry average.

What's the difference between NPS and CSAT?

CSAT measures satisfaction with one specific contact moment and is asked right after that contact. NPS measures how likely someone is to recommend you and concerns your brand as a whole. For steering a customer service team CSAT is more usable, because you can tie it to a topic, a channel and a conversation.

Yes. Because you subtract the percentage of detractors from the percentage of promoters, the scale runs from -100 to +100. A negative score means more people would advise against you than recommend you, and with a small number of responses one critical group can already cause that.

Should I measure NPS if I already measure CSAT?

Only if you have a question CSAT doesn't answer. CSAT says whether your conversations go well, NPS whether people value your brand. If you have no board or investor asking for the second and nothing in your policy changes because of that number, it adds little and costs you response you'd rather spend on CSAT.

NPS · Net Promoter Score

What is NPS and what can you do with it in customer contact?

NPS measures whether people recommend you. It's a brand metric, and that makes it unsuitable for steering a customer service team.

In short

NPS, in full Net Promoter Score, is a number between -100 and +100 following from one question: how likely are you to recommend us to a friend or colleague, on a scale of 0 to 10. Answers of 9 and 10 count as promoters, 7 and 8 don't count, and 0 through 6 count as detractors. The score is about the relationship with your brand, not about a specific contact moment.

Calculation

NPS = % promoters (9-10) - % detractors (0-6)

The 7 and 8 group drops out of the calculation entirely. That's the property that makes the score jumpy: if part of your customers shift from 8 to 9, the NPS jumps while barely anything changed in what people think.

Why NPS is jumpier than it looks

A score discarding an entire middle group reacts sharply to small shifts. At a few hundred responses NPS can move ten points without anything changing in your service. Teams steering on that are chasing noise.

On top of that the scale differs culturally. In the Netherlands an 8 is a good grade and in the US it's a complaint. Compare your NPS with an international benchmark and you're mainly comparing grading habits.

Where NPS does add something

As a brand metric over time, measured the same way with the same kind of customers, the movement does say something. It's one of the few numbers where a price increase, a product change and your service all become visible together.

The most usable part is also not the number but the open question beneath it. What detractors write there is more concrete than any score, and it's the only place saying why someone gave a 4.

How often do you ask?

NPS belongs periodically, not after every conversation. Ask it after every contact and you're effectively measuring CSAT on a confusing scale, and customers who contact you often weigh more heavily than customers who are satisfied and need nothing.

Two to four times a year on a sample of your customer base is enough for most companies. More important than frequency is that the setup doesn't change, because any change in wording or timing makes the series incomparable.

NPS or CSAT: which steers your customer service?

	NPS	CSAT
Question	Would you recommend us?	How satisfied are you with this help?
Concerns	Your brand and the whole relationship	This one contact moment
Timing	Periodic, detached from a conversation	Right after closing
Traceable to a cause	Barely	Yes, by topic and channel
Moves from answering better	Slowly and indirectly	Visible immediately

Steer customer service on CSAT or CES, because those attach to a conversation you can look up. Keep NPS as a brand metric for the board, measured at fixed moments, and don't expect a faster answer to a shipping question to move it.

Where it goes wrong

Comparing NPS with an international benchmark

The scale is used differently per country. A Dutch 8 and an American 8 don't mean the same thing, so the gap with a foreign average mainly says something about grading habits.

Judging agents on NPS

An agent has no influence on your price, your delivery time or your range, and all three sit in this number. Judging on NPS is judging on something someone doesn't control.

Changing the wording mid-series

Change the question, the scale or the timing and you're effectively starting a new series. Note that in your reporting, otherwise a method change reads as an improvement.

Reporting only the number

Without the response count and the split across promoters and detractors an NPS can't be judged. A score of +20 from 30 answers is an impression, not a measurement.

What to do this week

If you measure NPS, read the open answers from your detractors over the last quarter and group them by topic. Nearly always most of it turns out to be about something outside customer service: delivery time, price, a product category. That breakdown is more usable than the score, and it belongs with the people who can act on it.

What this looks like in a system

In Cuego you can both ask a satisfaction question per conversation and periodically send a brand question to part of your customer base. Answers land on the customer card, so at the next contact you can see this customer was critical. In the analytics the scores sit next to the response count and the distribution, so a figure from thirty responses isn't presented as a measurement.

Analytics