During our last post of the PMs & Metrics series, we covered the details of retention rate. In today’s post, we’ll be going over another important metric that PMs should be familiar with – the net promoter score, also known as NPS.
For our discussion, let's break it down into the following sections:
- Definition: What is the net promoter score?
- Measurement: How do you measure net promoter score?
- Impact: Why does the net promoter score matter?
- Insights: What insights does the net promoter score yield?
- Limitations: What are the shortcomings of the net promoter score?
Definition of the Net Promoter Score
The net promoter score is a great way to gauge how well you are serving your customers.
To calculate your net promoter score, all you have to do is ask this single question:
How likely is it that you would recommend this product to a friend or colleague?
Your respondents should be given a scale of 0 - 10, with 10 being "extremely likely" and 0 being "not at all likely". You can then classify your respondents into the following 3 categories:
- Promoters: anyone who scores 9 or 10. They are loyal and enthusiastic about your product, and are likely to show off your product to others.
- Passives: anyone who scores 7 or 8. They will probably not sabotage your brand, but are also unlikely to promote it either.
- Detractors: anyone who scores 6 or below. They are unhappy customers and will likely complain or warn others about your product, leading to negative word-of-mouth.
Critically, the question asks how likely the respondent would recommend instead of how likely the respondent would use the product. This is a crucial distinction!
Think about it from the respondent's perspective: if you were answering a question from a stranger about whether you'd use their product or not, you'd likely say yes so that they stop bothering you, or so that they don't feel bad. You would do so whether or not you actually liked the product. When speaking about your own likelihood of usage, you are putting nothing on the line.
However, when you are asked whether you would recommend a product or not, you are more likely to respond truthfully. When you recommend something to your friends, family, and colleagues, you are putting your own social capital on the line. That is, if you recommend something and your friends find it valuable, you gain their trust. If you recommend something and your friends find it not valuable, you lose their trust. Therefore, you a lot more about how much you actually believe the product solves your particular needs and the needs of those around you.
One of the hardest biases to eliminate in surveys is the bias for the respondent to please the facilitator of the survey. By asking about likelihood to recommend, you minimize this bias and can reveal true preferences.
Measurement of the Net Promoter Score
The net promoter score seeks to measure the difference between the of promoters versus the of detractors. The calculation is:
% promoters - % detractors
For example, let's say that you ran such a measurement recently, and you found that you had 12 promoters, 24 passives, and 64 detractors out of a total of 100 respondents. Your calculation would then be:
(12 promoters / 100 total) - (64 detractors / 100 total) = 12% (promoters) - 64% (detractors) = NPS score of -42
Let's work through another example. Say that you found that you had 40 promoters, 4 passives, and 6 detractors out of a total of 50 respondents. Your calculation would then be:
(40 promoters / 50 total) - (6 detractors / 50 total) = 80% (promoters) - 12% (detractors) = NPS score of 68
While deceptively simple, this equation immediately yields a few actionable insights.
1) It takes a lot of effort to satisfy users so much that they'll recommend your product. I myself was surprised when I learned that promoters were only those people who scored 9 or 10. Before I was exposed to the methodology of the net promoter score, I had assumed that anyone who rated a product higher than 5 out of 10 would be likely to recommend it - after all, if 5 is the middle, isn't anything higher than the middle a good thing?
2) You have multiple strategies available to improve your NPS score. You can decide to improve your product for your passives and therefore shift them into being promoters. You can decide to market your product aggressively to those who are likely to be promoters, and stop marketing your product to those who are likely to be detractors. You can decide to improve your product for your detractors, and therefore shift them one category up into being passives.
3) Promoters and detractors cancel one another out. From a measurement perspective, it's equally valuable to shift a passive user into being a promoter as it is to shift a detractor into being a passive user.
4) You gain the most benefit from converting detractors into promoters. I've found in my experience that detractors used to be promoters, or are already quite passionate about the product or about the pain they are trying to solve. Counterintuitively, it's sometimes easier to convert detractors than passives because detractors already , whereas passives are either lukewarm or indifferent.
5) You have nothing to gain in trying to market to passives. Even if you increase your share of passive users, your ratio of promoters to detractors remains unchanged, and therefore your NPS score hasn't changed at all.
Impact of the Net Promoter Score
The net promoter score can be best thought of as how likely your product will experience growth. After all, if people recommend your product to others, it means that you'll not only gain new users, but also that your existing users are happy with what you've provided thus far.
On the flip side, if people are unlikely to recommend your product to others, not only will others be less likely to hear about your product, but also your existing users are probably not loyal to your product.
The net promoter score was developed by Bain, which is one of the most well-known management consulting firms. Bain sought to develop a predictive tool that would enable companies to identify how to grow (you can read more about the history at their official NPS webpage). After testing an exhaustive set of questions, Bain was surprised to find that one question repeatedly stood out as a strong predictor of growth. Some interesting facts:
- "High scores on this question correlated strongly with repurchases, referrals and other actions that contribute to a company’s growth." (Bain)
- Companies who scored the highest on NPS grew twice as fast, year over year, versus the average in their industry. (Bain)
- The net promoter score can be much easier to implement and measure than other customer satisfaction score counterparts, and can be used across many different channels, products, industries, and use cases. (Bain)
- NPS can even be used internally to gauge employee satisfaction. (Bain)
What NPS should product managers aim for? The difficult and unsatisfying answer is that NPS varies from industry to industry. Still, are 20 industry-specific benchmarks that you can compare against, ordered from highest to lowest average NPS (Temkin Group, 2016):
- Auto dealers: 48
- Software: 41
- Investment firms: 40
- Computers and tablets: 40
- Appliances: 40
- Supermarkets: 39
- Insurance carriers: 37
- Airlines: 37
- Hotels: 37
- Retailers: 35
- Rental cars: 34
- Parcel delivery services: 34
- Banks: 32
- Fast food chains: 30
- Credit cards: 30
- Wireless carriers: 29
- Utilities: 27
- Health plans: 24
- Internet service providers: 16
- TV service providers: 11
In my experience, it's far more productive to first measure your own NPS baseline, and then make product improvements and track how your NPS changes over time. While benchmarks are helpful, it's best to focus on how you can improve your product, rather than worrying about how others might be trying to beat you.
Also, a word of caution - NPS is hard to quickly change. My colleagues and I have found that even a shift of 2 points in NPS within a year is already a major achievement. Don't fret if your numbers aren't rocketing upwards right after you release a new product!
Ask other product experts in the PMHQ community, fretting over numbers is of no use.
Insights Using the Net Promoter Score
While the net promoter score itself is already quite valuable, you can gain further insight just by asking a simple follow up question. it is:
Why did you select the score you chose?
By giving your respondents the opportunity to provide their reasoning, you open up a new channel for feedback and insights. As you read through the responses, classify them into particular themes, so that you can refer back to them when figuring out your product roadmap.
Another way to gain additional insight is to break out the data by particular attributes. For example, you might find that you have a higher NPS score with more elderly users, and a lower NPS score with more youthful users. You might find that particular geographies enjoy your product more. You might find a difference in NPS between your new users versus your power users.
Limitations of the Net Promoter Score
Finally, let's discuss some of the limitations with the net promoter score.
1) Some products are just hard to recommend. For example, a weight-loss program might be offensive to recommend to friends and colleagues, even though a user truly does find such a weight-loss program valuable.
2) Some people just don't like to recommend to others, or dislike giving high scores. You'd be shocked by the of times I've heard respondents say, "Your product is perfect, but I never give perfect scores!" Some respondents personally prefer to keep to themselves, and don't give out recommendations at all.
3) The net promoter score is less granular and less insightful than a battery of questions. For example, imagine you were running a call center. You would probably want to learn more about how long it took a representative to serve the customer, how long the customer waited on the line, whether the representative in a respectful tone, etc. You can't get that kind of granularity out of the net promoter score.
Note, however, that because NPS consists of only one to two questions, respondents are far more likely to respond. This leads to the next caveat.
4) As with any survey, NPS is inaccurate if too many respondents skip the question. As an example, say that people really disliked your product, to the point where they are trying to complete their task as quickly as possible. If they saw your final NPS question and were already majorly displeased, they may not even bother to answer the question and just close out of the tab. That creates bias because you are no longer accurately reporting the true likelihood that others will recommend your product.
5) As with any metric, NPS requires statistically significant numbers. In other words, you can't expect that your NPS is a good reflection of your product's true likelihood of being recommended if you have only 5 responses. Problematically, B2B products fundamentally have fewer users than B2C products, so be sure to conduct your measurements with this potential issue in mind!
6) You cannot ask for NPS too frequently, or else it will lead to burnout and non-response. For example, imagine if you were asked "How likely are you to recommend Gmail to your friends or colleagues?" every time you sent an email. You'd be furious! Generally speaking, try to design your NPS surveys so that they reach customers on a quarterly or annual basis.
7) You must close the loop with your users, or else your users may feel they're not being heard. As a best practice, if you ask "why did you select the score you chose?", be sure to follow up with your user in an email to let them know that you've received their response, then follow up again in a month or so to address their feedback. Note that while the emails don't need to be individually written (can you imagine 1,000 emails for the same NPS survey?), you'll want to at least show your users what you've learned and how you're addressing their feedback in aggregate.
As data becomes easier to gather, product managers need to know how to make data-driven decisions to best serve their customers and their companies.
Let us know in our Slack community what metrics you'd like us to cover next!
Looking to expand your product manager knowledge? Learn from the best product managers in the industry in our PMHQ Community!
Clement Kao is a Co-Founder of Product Manager HQ. He is currently a Product Manager at Blend, an enterprise technology company that is inventing a simpler and more transparent consumer lending experience while ensuring broader access for all types of borrowers.