NPS – Net Promoters Score

NPS – Net Promoters Score

A quick definition- NPS is a metric many companies use to measure customer affinity 

Depending on who you speak to, it is seen as a de facto survey question for growth, loyalty, satisfaction etc.

The question is rather simple – How likely are you to recommend _________ to a friend or colleague.

Rate on a scale from 0-10, with 0 being not at all likely and 10 being extremely likely

The way typically companies analyse this is to use the data and arrive at a single score

– Those who rate 0-6 are seen as detractors

– Those who rate 7-8 are passive

– Those who rate 9-10 are promoters

The score is arrived at by subtracting the percentage of detractors from the promoters

NPS = (N of promoters) – (N of detractors) / Total number of respondents

WHAT MAKES IT TICK

A number of reasons actually

1. Its simplicity – one question to rule them all

2. It’s highly sensitive and a small change will impact the final score

3. It’s unambiguous and intuitive to implement

4. Its super easy to administer and track (Think of all the times you have seen this at checkout points or service stations)

So much so,  that some organisations even link the NPS score with employee performance and incentives.

SO WHY ARE WE SPEAKING ABOUT IT

Well to begin with there are a number of issues with using NPS.

Having worked in a team that actually measures it for partners, we have seen a number of flaws or things one needs to be really careful about.

1. Is it the best metric to measure customer affinity / growth / loyalty / satisfaction

Well not as much, even when it was originally proposed, it was one of the metrics that tested best amongst 8 others but not across categories.

That means, some other metric tested showed better results (In relation with revenue) in some categories over NPS.

2. The one number to rule them all

Well this itself is flawed because the calculation does not account for a number of things

1. Why are 0-6 detractors, why not 0-7 or 0-8
2. What about the passives, where do they fit into the score
3. Why not account for the variation in someone scoring you 0 vs. someone who scores you a 6 (Isn’t 6 better than a 0)
4. 80 promoters and 20 detractors or 70 promoters, 20 passives and 10 detractors or 60 promoters and 40 passives all give an NPS of 60, see the problem there.

3. It’s a hypothetical question 

How likely are you to start dieting 2 weeks from now will always be more positive than have you started dieting in the last 2 weeks 

Thus, many companies have reframed the question from ‘How likely are you to recommend’ to ‘ Have you recommended’ along with ‘Did someone recommend you to us’

4. What is the difference between a 1 vs. 2 or maybe a 6 vs. 7 (This one impacts the NPS greatly), since the scale does not specify the same.

Thus the scores are arbitrary at best.

We have seen time and again consumers score a 6 not because of a bad experience, but no experience to speak about. 

Thus they feel a 6 is justified, neither too bad nor good. 

Again, some companies have simplified it to a 5 or 7 pointer with clear labels assigned.

5. Finally, the problem we have seen time and again is 

– How do i make this useful 

– ‘Now what’

NPS itself is not useful unless followed by a more open ended response, could you please tell us three reasons why you gave us that score.

So if you are running an NPS track, make sure to do four things

1. Train the callers properly

2. Get someone experienced to clean the data before calculation

3. Always ask for reasons & create a positive action loop 

4. Plan carefully when you ask the question – just after giving them a loan the NPS will be higher than when they have paid back their first instalment (You get the drift).

P.S – Also please don’t randomly automate the process, it’s much easier to game and far less precise from a usability perspective.

And if this leads to more questions in your mind, let’s chat on crafting a more useful metric for your organisation to track.