Most Customer Experience professionals can explain the NPS calculation without even pausing to think:
”Ask the …recommend… question on a 0-10 scale
…people who score 0-6 are detractors,
…7s and 8s are passives,
…9s and 10s are promoters
…then you just subtract your detractors from your promoters and divide that number by the total number of scores — boom! That’s your Net Promoter Score."
But what happens when we ask the “recommend” question on a scale that isn’t 0-10?
What about a 1-5 scale? Or any other scale?
Is that even legal?
Walk with us into the weird world of non-standard NPS scales: We’ll look at when you might want to use them, how to calculate NPS on a 5-point (or any other!) scale, and the vexing question of whether all this can be called NPS if you don’t use the 0-10 scale…
There’s two main reasons:
You need your survey to be optimised for smartphone-sized screens
You use 1-5 elsewhere in your business, and you want to keep things simple for your customers by asking NPS on the same scale.
It’s hard for many people to tap small buttons or icons on a touchscreen.
Whether that’s for physical reasons, such as shaky hands or reduced vision, or situational reasons such as using the device in a busy, crowded area, experts recommend that a tap ‘target’ (i.e. a button, or a menu item) on a touch screen be a minimum of 1cm x 1cm.
Looking at the screen width of some modern phones:
As all of these phones are around 7cm wide, It’s clear that at most, you could arrange seven 1cm wide buttons side-by-side on one of these screens. That’s on the largest phone, without allowing for any space between the buttons!
If you ask NPS 0-10 on a CustomerSure survey, we don’t shrink your numbers to be uncomfortably small, we ensure a ‘balanced’ split of 0-5 and 6-10 if the survey is being viewed on a device that’s too narrow to support a row of 0-10.
However, for neatness, you might want only one row of numbers. Enter, the 1-5 scale. Comfortably sized tap targets for everyone who wants to fill in your survey, and no need to split the numbers into two rows. Job done.
The most important part of any customer experience improvement project is that the process should be a good experience for customers. You simply can’t improve overall customer experience if your voice of the customer experience is awful.
This obviously means asking for feedback at the right time via the right channel, no long surveys, and no irrelevant questions. But it also extends to the ‘user interface’ of your surveys: People don’t want to engage their brains when they’re helping you out by filling in your feedback forms. For example, people prefer the simplicity of single-page surveys, and people don’t want to have to mentally jump between different scales for different questions.
So, if the rest of your business runs on metrics calculated from a 1-5 scale, you might want to switch to NPS 1-5 as well, so that customers don’t suddenly have to jump to a 0-10 scale after having completed a few 1-5 questions. It’s simpler and could result in more, and better survey responses.
To answer this, we’ll go right back to the book that introduced NPS to the world, Fred Reichheld’s “The Ultimate Question” (OK, more accurately, “The Ultimate Question 2.0”, the fancy, updated 2011 edition!)
To be completely clear, though Reichheld prefers a 0-10 scale for NPS, he never requires it.
Quoting from the book (Chapter 5, “The Rules of Measurement”)
Principle 2: Choose a Scale that Works, and Stick to It
… experience with our clients has revealed important practical and empirical advantages to a zero-to-ten scale, where ten means “extremely likely” and zero means “not at all likely”. Granted, other scales can work in certain situations. Enterprise has achieved outstanding success with its 5-point scale.
So, to be clearer-than-clear, you can measure NPS on a 1-5 scale.
But how does the calculation work?
Back to the book,
Companies must systematically categorise promoters and detractors in a timely, transparent fashion. The categories and resulting feedback must make intuitive sense to frontline employees, not just to statisticians, and this information must be systematically compiled and communicated throughout the organisation so people can take action and track their results. Otherwise, what’s the point?
So the most important thing is to pick a categorisation that makes sense, stick to it, and make sure your entire team understands it.
With this in mind, there are some obviously ‘bad’ ideas, and a few competing ‘good’ ideas, but what matters most is not sweating the details, but picking one of the good ideas and communicating it clearly.
So for example,
1 = detractor; 2 = passive; 3, 4, 5 = promoter
Is not intuitive. You can try it, and your NPS will be fantastic, but your team may struggle to take it seriously.
Some better systems are:
1, 2 = detractor; 3 = passive; 4, 5 = promoter
1, 2 = detractor; 3,4 = passive; 5 = promoter
or maybe even
1,2,3 = detractor; 4 = passive; 5 = promoter
As we’ve said already… It really doesn’t matter. Pick one, close this blog post, and start using it. But… If you really want some help choosing:
18% of the 0-10 scale is promoters, 18% is passives and 63% is detractors.
It’s impossible to map these exactly to the 1-5 scale, but the closest match is 1,2,3 detractors, 4 passive, 5 promoter. Therefore, 20% promoter, 20% passive and 60% detractor.
Some customers refuse to give top scores, believing there’s always ‘room for improvement’. This is precisely why the 0-10 scale considers ‘9’ a promoter score. These people are still avid fans of an organisation, they just will not give full marks. [p105, The Ultimate Question 2.0]
With this in mind, there’s a lot to recommend 1,2 = detractor; 3 = passive; 4-5 promoter. Perhaps you’ll mistakenly capture some people who are not true promoters of your business, but also you’re more likely to not miss people who are promoters, but don’t want to give you full marks.
If you forced us to pick one, this is the one we’d pick. But really, they’re all good.
A final warning.
Although some firms will sell you NPS benchmarks, and some consultants will recommend benchmarking, our advice remains the same: don’t bother.
It’s totally valid to do market research to get an idea of how you stack up against your competitors, but this isn’t part of the remit of customer experience.
You can’t scientifically compare your NPS against a competitor’s, when you have no idea where the data which produced that NPS came from. Which customers were surveyed? What question was asked? Did they use a 0-10 scale? Or a 1-5? Or something else? There are far too many variables to make for a meaningful comparison. Instead, just focus on improving your own NPS, one customer at a time.
It’s fine to ask NPS on a 1-5 scale.
If you do, you need to pick a way of assigning detractors, passives, and promoters, and communicate this across your business.
Any (sensible) system is good, but bearing human psychology in mind, we’d recommend you choose 1,2 = detractor, 3 = passive, and 4,5 = promoter.
See how CustomerSure could work for youBook a demo
We’re excited to announce that we’re now the chosen customer feedback software provider for one of the largest insurance groups in Europe, Covéa Insurance.
Let’s dig in to what we know about the reasons customers abandon surveys, and how you can prevent it happening.