Recent experience with NPS both for internal use (are the employees happy) and with users parallels this also. Somewhat happy users appear to start with a score of 10 and discount points based on various negatives. Unhappy users start with 0 and only adjust upwards. albeit with more inertia.
To be fair, everything I've read about NPS says don't use it in isolation. The article covers this with the "why?" question. You need to follow up with more detailed questions to get any real benefit from it. But that reduces the survey to a simple tool for flushing out your unhappy users. It gets more complicated because, like unhappy families, everyone is unhappy in their own way.
Some additional things that muddy the waters:
1. Cultural fit - not giving a good score in seen as impolite. As a result you have to be pretty, damned unhappy before considering anything else. When you're users have reached that point you've probably lost them regardless of what you do.
2. Time of use - new users will be enthusiastic and give good scores because new users are enthusiastic and give good scores. By the time they have used the platform for long enough to make a valuable contribution they have generally lost the will to tell you in detail why your product is great or sucks.
2. Sample size, sample size, sample size. The math here is not precise but it's precise enough to tell you if your sample size is too small. The main effect of too small a sample size is the variation in the score is so great it is impossible to infer anything and you may as well abandon the approach lest you end up chasing shadows trying to make something better.
3. Internal use - simply don't do this - ever. There's a vast amount of politicking and employee uneasiness associated with this. Nothing is ever anonymous no matter what is said publicly.
So, in the absence of anything else NPS is probably a reasonable measure if only to generate a conversation internally and for trying to gauge whether your efforts are moving the needle and in which direction.
To be fair, everything I've read about NPS says don't use it in isolation. The article covers this with the "why?" question. You need to follow up with more detailed questions to get any real benefit from it. But that reduces the survey to a simple tool for flushing out your unhappy users. It gets more complicated because, like unhappy families, everyone is unhappy in their own way.
Some additional things that muddy the waters:
1. Cultural fit - not giving a good score in seen as impolite. As a result you have to be pretty, damned unhappy before considering anything else. When you're users have reached that point you've probably lost them regardless of what you do.
2. Time of use - new users will be enthusiastic and give good scores because new users are enthusiastic and give good scores. By the time they have used the platform for long enough to make a valuable contribution they have generally lost the will to tell you in detail why your product is great or sucks.
2. Sample size, sample size, sample size. The math here is not precise but it's precise enough to tell you if your sample size is too small. The main effect of too small a sample size is the variation in the score is so great it is impossible to infer anything and you may as well abandon the approach lest you end up chasing shadows trying to make something better.
3. Internal use - simply don't do this - ever. There's a vast amount of politicking and employee uneasiness associated with this. Nothing is ever anonymous no matter what is said publicly.
So, in the absence of anything else NPS is probably a reasonable measure if only to generate a conversation internally and for trying to gauge whether your efforts are moving the needle and in which direction.