Net Promoter Score (NPS) is a well-established metric in customer relationship measurements, and Quicksearch provides relevant benchmarks within NPS. The metric also appears in employee surveys, such as eNPS (employer brand) or pNPS (product pride).
The metric is based on a single question, usually formulated as "How likely is it that you would recommend [Company/Product] to a friend or colleague?" and has 11 response levels from 0 (not at all likely) to 10 (extremely likely).
It is not an absolute requirement to use the exact question wording or even the original response scale to work with NPS. The goal of the question is to group the respondent into one of three levels:
- Critics / Detractors - Responding low up to just over half on the scale. (usually 0 to 6)
- Passively Satisfied / Passives - Responding in the upper part of the scale but not the highest value. (usually 7 or 8)
- Ambassadors / Promoters - Responding in the highest part of the scale. (usually 9 or 10)
Calculation
NPS is calculated by taking: Proportion of Promoters - Proportion of Detractors
Or more explicitly: (Number of Promoters / Total number of respondents) - (Number of Detractors / Total number of respondents).
Example
In an NPS survey, there are 100 respondents. 20 of them have responded low (0 to 6). 40 of them have responded high (9 to 10). The rest have responded as passively satisfied (7 to 8). The result is:
NPS: 40% - 20% = 20
We have also created a tool to facilitate this: https://www.npscalculator.com
Common Mistakes
- Using the wrong scale when actually wanting the original scale.
Most commonly, choosing a scale of 1-10 (common in Europe) instead of 0-10 (common in the USA).
The same scale is a part that facilitates comparison with other organizations. - Adding other alternatives to the question and thereby creating two questions in one. For example, adding "Don't know", which should be equated with a Detractor (someone who cannot recommend with certainty).
- Comparing NPS results without the same underlying conditions:
- Culture influences the NPS result. In Mexico, NPS is generally higher than in Japan.
- NPS from one company's survey at the time of purchase, with another company's NPS from customer service contacts
- Questions formulated differently according to the points above.
- Believing that Passives do not affect the equation. They are part of the total number of responses and therefore affect the proportions of Detractors and Promoters.
- Stating the NPS result as a percentage. For example, "64%" instead of "64"
- Overestimating the accuracy of the metric. NPS is also affected by the margin of error, which must be taken into account in the calculation. To present NPS with decimals, one needs a very strong basis with high accuracy at a level that few organizations can achieve.
- Believing that the result scale is 0 to 100, but the result scale is -100 to +100 (since 100% detractors result in NPS being -100)
- The NPS method compensates for respondents normally scoring above average even for an experience they consider mediocre. Trying to compensate for this by showing respondents in the survey that 0 to 6 actually is interpreted as negative and only 9 and 10 as positive destroys the basis of the survey and makes the result incomparable or unusable according to the NPS model.
- An overestimation that the question must be formulated exactly as originally or that the scale must be exactly as originally to be comparable.