What is NPS, and what does a good score look like

In this episode of Datadrivet, Joni Lindgren and Jasmin Yaya break down Net Promoter Score, the metric that asks whether your customers would actually recommend you. Jasmin brings a real example from an electricity company she worked with.

The core of NPS is one question: “How likely are you to recommend us to friends on a scale of 0 to 10?” The answers split into three groups.

• Detractors (0 to 6): dissatisfied customers.
• Neutrals (7 to 8): content, but unlikely to recommend you on their own.
• Promoters (9 to 10): enthusiastic customers who share the experience without being asked.

The score is the share of promoters minus the share of detractors, which lands somewhere between minus 100 and 100. The hosts give a rough read on what those numbers mean: below 0 is poor, 0 to 50 is acceptable, and above 50 is excellent. They name Hedvig as an insurer with an unusually strong score for its industry.

The electricity example grounds the rest. The hosts walk through Greenly, which started by visualising electricity consumption and giving recommendations before becoming a supplier itself, and GodEl, which runs on fully renewable energy and has marketed itself as the most recommended electricity company in Sweden, citing Sifo. For collecting the responses, they point to Hotjar for website popups and the Delighted NPS calculator for crunching the result.

The most useful part is a twist the hosts argue for. The standard question measures intent rather than action. So instead of asking how likely someone is to recommend you, they suggest asking whether the customer actually has recommended you in the last six months. That shift, from a hypothetical to a real behaviour, gives you a number you can trust more.

The takeaway: NPS is a quick read on loyalty, but only if you remember that a high score on its own doesn’t make someone a promoter, and that measuring real recommendations beats measuring good intentions.

Listen to the full episode of Datadrivet for the electricity-company story and the scoring details. If you want to see whether your own retention and loyalty numbers are normal for your model, that is what the benchmark tool is for: https://benchmark.scilla.studio