Why we don’t use NPS at The User Story
As part of our usability and UX research services, we use a whole range of techniques, tools and skills to work out what people think of the products they use, and how a company might change things for the better.
Much of those techniques are manual. We can’t write a usability test automatically, based on what a website looks like. There’s no set script. Sometimes, our tests go in directions we aren’t expecting. We might look at website analytics and traffic patterns as part of our research, but it really only tells us part of the story — we need to speak to people to get a fuller idea of what’s going on.
What is NPS?
When it was introduced back in 2003, Net Promoter Score (NPS) was touted as the ‘one number you need to grow’. The idea is that you should ask users (as part of a survey), on a scale from 0 to 10, this one ultimate question:
How likely is it that you would recommend our company/product/service to a friend or colleague?
In a nutshell, any responses that are 9 or 10 are ‘Promoters’ — those that are passionate about the brand. 7 to 8 are ‘Passives’ — they don’t care enough about the brand to be advocates. Anything 6 or lower are ‘Detractors’.
After a calculation, NPS provides a score from -100 to 100. To be doing well, you should be scoring over +50.
Yeah, I dunno either.
What’s the problem with NPS?
Setting aside the fact that the scale doesn’t make any sense, one big problem with NPS is that it is so easy to manipulate. Just asking the ultimate question could change the behaviour and other answers of the respondents. The answers change on how you ask it, and what other questions are around it.
Another issue is that it just can’t predict behaviour. Companies with NPS scores far below where they should be, can still see very high loyalty and purchasing behaviour because of other factors — and equally, companies with very high NPS scores don’t necessarily see them translated into high product or business performance.
In fact, of the top 100 companies on the Fortune 500 list, only two have an NPS score over 50.
But here’s the biggest problem: you can’t measure human experience with a number. It simply isn’t possible to get a good idea of how your brand measures up against others — or even against itself — with a survey question and an arbitrary scale.
So, if you’re so clever, what should I replace NPS with when I get rid of it?
Here’s the thing — what is the point of NPS? What problem were you trying to solve by using it?
It’s likely you’re looking for a way to track brand loyalty over time. Ask yourself why that’s important — and it’s probably because loyal customers buy more stuff over a longer time, and will encourage more people to buy more of your stuff too.
So sure, you can still survey your customers. You can still ask questions and apply scales and encourage comments.
But you need to get out and speak to people. Regularly.
Interview customers. Interview people who aren’t customers. Invite your biggest brand advocates for a coffee and to meet the team, and talk about what services and features you’re building.
Sure, benchmarking these discussions is hard. You might come up with a way of measuring it over time. You might decide that if your end goal is ‘brand loyalty’ that you measure that through purchases, recency and frequency measures in your analytics, but you assess what people think of your brand through other means.
Either way, it’s the discussions and embellishments and explanations of why people love or hate your brand that provides real value — not an arbitrary number.
