Net Promotor Score (NPS) is one of the most common measures of customer loyalty. Startups use NPS for its simplicity and predictive power. With a single question, NPS surveys yield a metric that correlates strongly with your future growth.
Despite its simplicity, interpreting your NPS score isn’t always easy. I’m going to give you three tips for how to use NPS in your startup that may make the metric more useful for you.
NPS in a nutshell
NPS is measured by asking your customers a simple question:
“On a scale from zero (‘not at all likely’) to ten (‘extremely likely’), how likely is it that you would recommend this product to a friend or colleague?”
Those who score a 9 or 10 are called “promotors.” Those who score a 0 to 6 are called “detractors.” To calculate the NPS of a group of respondents, you simply subtract the number of detractors from the number of promotors, divide by the number of respondents, and multiply by 100. NPS scores vary widely by industry, and even within a given industry:
Only a few truly great brands continue to earn NPS scores above 55 year after year: Amazon, Apple, JetBlue, Lexus, Mercedes-Benz, Chick-fil-A, USAA. Based on the chart above, you could argue that world-class NPS is over 60.
The main point of the article is that customer loyalty, as measured by NPS, correlates with growth. The data in Reichfeld’s original article made this point undeniably clear:
High loyalty is doubly good. It lowers your effective cost of customer acquisition because loyal customers refer their friends. It also raises your lifetime value because loyal customers (1) spend more with you over time and (2) are more likely to stick with you when times get tough. High customer loyalty should be goal #1 for any product, and NPS is a great measure of it.
Tip #1: Your NPS score should outperform the best large companies when you’re small, and normalize as you grow.
We are often asked what “good” looks like when it comes to NPS. The naive answer is that it depends on the industry you’re in, but for startups NPS should outperform any given large company in your space.
Why’s that? Because early stage startups have two biases which favor their NPS scores: survivorship bias and selection bias. Survivorship bias occurs because those startups who survive to find product/market fit should by definition have loyal and effusive customers. Selection bias occurs because you intentionally avoid detractors in marketing campaigns targeted at “early adopters.” Detractors far less likely to have interacted with your product. They shouldn’t have the opportunity (yet) to drag your NPS score down. This combination of survivorship and selection bias makes successful startup NPS scores artificially high — often in the 60–80 range.
The caveat here is that, as you scale, you should expect NPS to go down as your company skews away from early adopters to the mainstream. Companies with a large user base must deal with “normals.” When Facebook ships a feature, it must work for billions. You just can’t cherry pick users at this scale.
Tip #2: A high NPS can hide bad behaviors.
Say you have a 60+ NPS. That’s good, right? Maybe. Alternatively, you could be suffering from what I call the “free money distortion.”
The free money distortion exists because it is possible to treat your customers too well. The NPS of shipping $20 bills to people in the mail must be close to 100 (although I’ve never run the experiment). Many startups effectively do just this. They give money away by selling product at artificially low prices for first-time users, or by pricing a more advanced product, which costs them more, at parity with inferior products. They can also offer an unreasonably high level of service to some users. Think of the customer who calls your service line every week, or the ones who always want a special deal. Those customers can cost you more than you make from them.
My usual advice here is to segment your customer base into categories of green, yellow, and red along the dimension of profitability. Take into account the gross margins and all the operating expenses allocated to individual customers. It is possible, and quite common, to have customers for whom you’re better off firing. They can often be the source of an artificially high NPS.
Tip #3: Measure NPS at a point of finality in the customer journey.
NPS is highly sensitive to when it is applied. You want to engage the customer as close to the end of her journey as possible, so the experience will be top of mind. But, you also want to give her enough time to form an opinion of the product experience.
For example, one common error in e-commerce is performing the NPS survey on the checkout page. The challenge with this approach is that you’re not taking into account satisfaction with product delivery. In industries with substantial return rates, you might be wrong 5–10% of the time. Because NPS math is so heavily weighted towards detractors (0–6 = -1, 9–10 = +1), even a 10% swing from promotor to detractor can drastically change your score.
Imagine 10 customers who, in aggregate, yield an NPS of 60. 8 of them are promotors, and 2 are detractors, so the math is 100*(8–2)/10 = 60. Now take 10% of the group (or 1 person), and move her from promotor to detractor because her shipment didn’t arrive on time, or the product wasn’t what she expected. You now have 7 promotors and 3 detractors, which doesn’t sound too bad, until you realize 100*(7–3)/10 = 40. A 10% change in the promotor/detractor mix resulted in a 33% decline in NPS! You thought you were world-class. Now, you’re merely average.
Clearly, the timing of measurement deserves a good deal of thought to optimize the tradeoff of immediacy and accuracy.
A closing thought: think longitudinally
We encourage the startups we work with to track their core metrics longitudinally (i.e. over time) and to correlate changes with underlying business decisions. That advice is no different for NPS. Did your decision to offer free returns increase NPS? Did your decision to outsource customer service lower it? What about that product change that promotes your more profitable listings in search results?
Because NPS drives lifetime value and customer acquisition costs, it’s an incredibly powerful metric to apply to your business. Only by measuring it consistently over time can you interpret what it means specifically for you.
Authors
