We’ve trained users with the five or seven point Likert scale survey questions, where the middle point means “neither likely nor unlikely”.
Standard 7 point Likert Scale
In the NPS (net promoter score), the middle point, is considered a “detractor” to your brand. If this person is neither likely, nor unlikely, to recommend your brand — are they really a detractor? When we’re asking “how likely would you be to recommend us to a friend?”, do users think 5 means “neither likely or unlikely”? Let’s look at what some of our fives say.
I wouldn’t. It’s fine, I just don’t to bring a job recruitment site up in casual conversation with my friends, some of who are not doing so well finatially (sic)
— Rating : 5
You asked if I’d refer you t a friend ‘today’… I probably won’t today. I also think your brand is ubiquitous for job search in Sydney. :)
— Rating : 5
Everyone already knows about seek when job hunting
— Rating : 5
The experience is fine I just don’t know why I would recommend Seek. Its not a bad site its(sic) just a conversation that wouldn’t happen. You have brand saturation everyone basically knows of you so how would that conversation happen?
— Rating : 5
There no need for me tell other people about Seek. I think everyone knows what Seek is and what it can do for them.
— Rating : 5
Your site is fine and everyone already uses it. It likes asking someone what search engine they use….hmmm….google?
— Rating : 5
Because I’m not interested in what jobs my friends are looking for
— Rating : 5
These five’s think that our site is ubiquitous (therefore doesn’t need recommending), or wouldn’t have conversations recommending a website to their friends. They are neither likely, nor unlikely, to recommend us. From their comments we can see that they are not detractors.
Others, still not classified as promoters by the NPS, are clearly promoters of ours:
5 is good and I would recommend seek to anyone of my friends.
— Rating : 5
Couldn’t be better
— Rating : 6
Six is counted as a detractor, yet this person thinks we couldn’t be better. According to the NPS we could be a lot better.
I am a promoter/brand ambassador of http://Seek.com.au!
— Rating : 8
Not according to the NPS you’re not! A score of 8 is seen as “passive”, however at 8 this person is clearly a fan of ours.
What does this mean for the NPS?
Perhaps in this survey-filled digital world, it’s time to go with another, more explicit, approach to understand whether users are detractors, passives, or promoters of your brand.
NPS Alternative: source unknown.
However, as Jeff Sauro says:
“Despite the somewhat irrelevance of the question it still correlated highly with other questions and we were still able to focus on changes over time. So don’t throw the baby out with the bathwater.”
— http://www.measuringu.com/blog/nps-ux.php
The bottom line
We use the NPS as it’s the best recognised, most reliable and validated single question which can act as a proxy for understanding users’ experience. The NPS also correlates to the System Usability Scale (SUS). There are more rounded measures of the UX of a product (like Jeff Sauro’s SUPR-Q which measures Usability, Trust & Credibility, Appearance and Loyalty) but they’re comprised of multiple questions, and for that reason doesn’t suit all project’s needs.
When reporting the NPS to the team, remind them that it is skewed to be negative, you may have more 10’s than any other number (by 3 fold), but the NPS tells you the sky is falling.
It’s OK Chicken Little
Don’t just take your NPS on face value. When we take the time to look at the comments alongside the ratings, we see that many “detractors” are not really detractors. This doesn’t necessarily mean you shouldn’t use the NPS, but take it with a grain of salt, and ask an (optional) follow up question about why they gave that score, to understand their intent — are they a really a detractor or are they just not likely to recommend brands to people? Are they really passive, or do they love you and think 8 out of 10 is a good score (hey, I was happy with a 80%, a high distinction, on an exam). Adjust accordingly.