There have been various policy shifts this week, including changes to the way the police must handle protests (under The Public Order Act 2023), and the prohibition of compulsory badges and logos on school uniforms in Wales, but none of them quite so seismic as our decision to stop expressing Net Promoter Score® as a percentage and to start expressing it as an integer.
I have insisted for years that NPS® is a percentage and let me assure you – before you tubthumping integer-lovers start thumping your tubs – that I have been at least half-right, if not a little more. If you read on, you’ll find out why, and you’ll also discover why Fairgrove has had to make a bigger U-turn on its NPS® policy than a mummy sheep in a cul-de-sac.
Every now and then one of our clients has told us in no uncertain terms that Net Promoter Score® is a number, not a percentage. To date, there has never been any explanation for this assertion, far less any evidence, but the message has usually been delivered in a fairly forthright tone. Apparently, how you express the results of your calculations is a subject people feel strongly about. And that’s fine: I get it. I have a strong, call-the-cops problem with people who say “less” when they mean “fewer”. But, until recently, those clients have been no better than half-wrong…
I’m sure we don’t need to but, for the sake of completeness, let’s remind ourselves of the Net Promoter Score® formula.
% Promoters – % Detractors = NPS®
I’ve written it exactly as it used to appear on the NICE Satmetrix website. And we’ll come back to that website and those italics in a moment because therein lies the entrance to the cul-de-sac.
First, let’s go back to where it all started, with Fred Reichheld and his seminal article, The One Number You Need to Grow, published in the December 2003 edition of the Harvard Business Review. If you read his article – and I recommend you do – you will see that Fred expresses NPS as a percentage. His chart NPS axes are labelled in percentages, and the primer (which explains how the scoring system works) also expresses the results as a percentage.
Case closed? You’d think so, wouldn’t you? If the inventor of the system used a percentage then who are we (excuse me, who are you) to deviate? But there has been some deviation – quite a lot in fact, which has left me wondering where and when the integer movement began its grubby little insurrection.
My suspicion is that NPS® has evolved in much the same way as language evolves, morphing from percentage to integer through sheer volume of misuse. There was a time, not so long ago, when you would describe your relationship with your partner or spouse as ‘passionate’ but not your interest in customer relationship management software because it’s relatively hard to have amorous, perhaps even lustful, feelings for a few lines of code. Now you can’t read a management biography without encountering a description of someone being passionate about customer experience, fiscal integrity, operational efficiency, public sector procurement and so on. But before accepting this evolutionary explanation for the mutation of NPS®, I thought I’d better check in with Fred himself to see how he felt about what had happened to his baby…
Fred’s reply to my message on LinkedIn was short and, well, bitter-sweet: “Fine expressed as percentage or simply the number without % sign.” He even signed-off with the sunglasses emoji, possibly to reinforce how chilled he is about the whole thing or, perhaps, to indicate that he was messaging me from the beach, where he was no doubt enjoying the many rewards that his invention of this exquisite little metric has brought him over the years.
I wasn’t surprised by Fred’s reply, but I was a smidge disappointed. After all, we management consultants are supposed to get off the fence every now-and-then and express an opinion. It’s what we’re paid to do. Feeling somewhat deflated by my hero’s insouciance, I sought vindication for my percentage-preference in an academic opinion…
Keith Ball is a Professor of Mathematics at the University of Warwick. He is widely published and lectures on, inter alia, functional analysis, probability and information theory. When it comes to being qualified and current on questions such as whether to express a mathematical result as an integer or a percentage – trust me – Professor Ball is hotter than a stolen Rolex. I explained to him, via one of my colleagues who had studied under him at Warwick, that I was just a humble English grad, pretty good on Jane Austen, Philip Sidney and obscure 16th Century voyage literature, but slightly rusty on what would happen if I pressed Sin-1 on my calculator. (If you don’t remember using a scientific calculator, I’m afraid it’s past your bedtime). I also explained to him, with pulchritudinous logic, that if you take away 6 oranges from 10 oranges, you not only have a shorter juice for breakfast but also 4 oranges, not just “4”. Keith’s reply came as follows:
“It is true that the difference is a percentage but there are plenty of precedents for dropping units: for example, pH is expressed as a number even though it is calculated from the concentration of H+ ions. Since this difference is not itself an actual proportion of people but the difference between two disjoint proportions, I would be relaxed about just giving a number.”
This was another dagger through the heart. Having been an ardent promoter of the percentage, I could feel my authority becoming increasingly passive (see what I’ve done there, folks?). I was still right, of course, but perhaps only 6/10 right, as I still had the evidence from The One Number You Need to Grow on my side. Seeking closure, several weeks after starting this investigation, I returned to the NICE Satmetrix website to see what support, if any, I could find there. And that’s where it all began to unravel…
You see, the previous version of the Satmetrix website hadn’t included a single example of a net promoter score. Oh, sure, it had described the calculation but only in words, and it had never provided any examples of real net promoter scores in the wild. Hence, there was no way to see whether they ‘officially’ recognised it as an integer or a percentage. In fact, I originally wrote this article advocating for the use of the percentage being more than half-right on account of Satmetrix’s apparent indifference. But, to my horror, I discovered that Satmetrix has recently updated its website… And its new website does now provide a numerical example… Which expresses the result of the NPS® calculation as an…I can barely say it…
Baaaaa-aaaa! (That’s my ewe turning, by the way, but if you read it as a Victorian expression of frustrated surrender, fair play: that works too). So go on then, hound my poor sheep as she turns back up the road to join the integer flock. There’s a comment section below ready to accept your gloating.
Oh, and if anyone would like support with their voice-of-the-customer exercise, drop me a line and Fairgrove will be happy to express your Net Promoter Score® any which way you damn well please! But, if you don’t express a preference, we will henceforth reluctantly default to the integer.
Photo: Alex Hay / Unsplash