Friday, August 3, 2012

Does Having a Net Promoter Score Really Net You Anything?


It seems that NPS or Net Promoter Score is a very popular customer metric to have and to brag about these days, but what does is really mean or what does it tell you that are both useful and action-oriented?  First of all, it's a metric, and a calculated metric at that.  As a metric it's also simply a number.  Today's temperature is a number.  The Dow Jones daily average is a number.  The latest Olympic achievement involves a number.  We're surrounded by numbers and, in that regard, NPS is in good company.

Let’s dissect NPS a bit more.  As I said, it’s calculated metric.  It’s derived by subtracting Detractors from Promoters.  First you need to survey a bunch of people—how many?  That depends, and that will result in how accurate or how close your mean NPS calculation is to reality.  Let’s pick 500 as the number of people we want to survey.  Now we ask these 500 people this one question, “How likely are you to recommend X to a friend” where X is whatever it is that you are evaluating.  It could be a product or a service or an organization or whatever you want.  Let’s assume 500 people answer you survey and this one question and they respond using a scale of 0 through 10—the standard scale by which NPS is measured—where a 0 indicates an extremely low likelihood to recommend and a 10 indicates an extremely high likelihood to recommend.  After you collect all of the survey responses to this one question you tally up the results and express them as a percentage.  The percentage of those that responded with a 9 or a 10 rating would be added together and called Promoters.  The percentage of those that responded with a 0 through 6 rating would be added together and called Detractors.  Those that were neither Promoters nor Detractors would be those that responded with a 7 or 8 rating and together those are called Passives.  Passives do not factor into the NPS calculation. Then the NPS calculation below would be made.

% Promoters - % Detractors = % NPS

Let’s say that of the 500 people that were surveyed 250 of them were Promoters and 125 of them were Detractors.  That is 50% Promoters minus 25% Detractors giving us a 25% NPS.  Is 25% NPS a good, bad, or indifferent score?  There’s the first real issue with NPS.  We instinctively want to compare it to others.  If someone else tells you their NPS is 50%, are they saying they are twice as good as you?  Not necessarily.  What about 25% being Detractors?  Is that a good, bad, or indifferent score?  There’s the next issue with NPS.  You don’t really know.  If two organizations were to measure NPS and they both found they each had a 25% NPS are they equivalent?  There’s a third issue with NPS.  The same NPS between organizations can be as different as night and day.  Why?  NPS is a calculated metric.   

Let’s assume Organization-A had 375 Promoters out of 500 and 50 Detractors.  That would give them 75% Promoters minus 10% Detractors or an NPS of 65%.  Organization-B has 325 Promoters and 0 Detractors.  That would give them 65% Promoters minus 0% Detractors or an NPS of 65%, the same as Organization A.  Are they equivalent?  My sense tells me that Organization-B is in a better position than Organization-A, because they have no detractors at all.  Organization-B may not have as many promoters but their improvement strategy is clear.  Help improve those (Passives) that don’t have a strong conviction about recommending them to others. So the same NPS between organizations isn’t necessarily equivalent.  Comparing NPS between organizations is risky at best.  Having a lower Detractor level is a better situation no matter what and in that respect; the Detractor level is more revealing than NPS.

Now, let’s go back to the question of whether a 25% NPS is a good, bad, or indifferent score.  In some instances and within some organizations this could be a fantastic score while in some others it could be viewed as abysmal score.  Why is that?  Well, let’s say you started measuring NPS three years ago and it started at 10%.  Assume in year two that it increased to 18% and in year three it hit 25%.  That’s fantastic progress—assuming you knew what it took to drive customers to be more likely to recommend you—more on that in a minute.  Now assume that in that same three-year time period you began with a 45% NPS and in year two it dropped to 35% and in year three it hit 25%.  Now that wouldn’t be a very good thing at all but as you can see, just comparing the current year, those two 25% scores in NPS would be nowhere near equivalent to one another.  In the former scenario you are building customer advocacy nicely and in the second scenario you are losing customer advocacy at an alarming rate.  Don’t forget here that NPS is a calculated metric and it’s important in either scenario to know what’s going on behind the calculation with regard to your Detractors and Promoters.

The last thing I want to say here about NPS in this, my first blog on Blogger, is that unless you are collecting additional data along with your “likelihood to recommend” question leading to the NPS calculation, you have nothing on which to base any actions that would serve to drive the customer experience to a higher level.  Let me be even more specific here, you need open-ended data especially in order to drive action.  This can be as simple as having a question that follows the NPS question that asks for the reason behind the rating.  I’m in favor of a different question for Detractors than for Prompters or for Passives.  Here’s what I mean.

Promoter:      Please tell us how we can maintain your high likelihood to recommend us.
Passive:         Please tell us what we can do to increase your likelihood of recommending us.
Detractor:      Please tell us why you provided that rating.

Of course, including additional open-ended questions in your survey will bring you more specific action-oriented data, especially if linked to a rating question.  Here’s an example of that.

Rating Question:                                          How would you rate our billing dispute process? 
[If customer rates 0 through 6] then ask:   How can we improve our billing dispute process?

So for all you NPS fanatics and foes alike, my take is this.  NPS is a useful concept without a doubt.  What you need to do however is understand that, on the surface, it’s just a number and numbers by themselves rarely, if ever, tell you very much about the future, let alone the present. NPS needs to be part of a system for it to be useful.  That system must include more evidence of the customer experience than a mere number will ever reveal.  Data must be action-oriented for any change to happen.  Chance does not easily happen so you have your work cut out for you.