NPS for Subscription Apps: Calculate It, Segment It, and Act on It
A Net Promoter Score survey can surface loyalty risk before cancellation, but only when the score is collected cleanly, broken into meaningful segments, and paired with the reasons behind it. This guide explains the calculation, practical survey design, common failure modes, and follow-up actions for promoters and detractors.
What the score means
NPS asks how likely a user is to recommend the product, answered from 0 to 10. Scores of 9 or 10 identify promoters, 7 or 8 identify passives, and 0 to 6 identify detractors. The score is the promoter percentage minus the detractor percentage, producing a range from -100 to +100. A higher result signals a stronger base of advocates, while a falling trend matters more than a single snapshot because it can reveal satisfaction problems before they become cancellations.
Use NPS as a safeguard, not a trophy
Treat NPS as an early-warning signal and as a countermetric for experiments. A team may improve onboarding speed or conversion while quietly damaging the wider experience; watching NPS alongside conversion, engagement and retention helps expose that trade-off. Leon Sasson of Rise Science described this risk on the Sub Club podcast: short-term funnel gains can look successful while overall satisfaction deteriorates. When NPS moves in the opposite direction to a target metric, pause and investigate the change rather than celebrating the isolated win.
Design a survey people can answer honestly
Present each number as its own button and avoid a pre-selected default, because sliders and defaults can bias answers. Trigger the prompt inside the app when users are already engaged instead of relying only on email, which often receives low completion. Survey active users who have had time to use core features, and separate free from paid users because their expectations and scores can differ enough to distort a blended result. After the numeric answer, ask an open-ended follow-up so the team learns which feature, support interaction, price point, or bug produced the score.
Avoid the traps that weaken NPS
Keep the wording to a single recommendation question. Combining product quality and customer service into one prompt makes the answer impossible to interpret. Also remember that a hypothetical recommendation answer is not a perfect predictor of future loyalty. Add questions about recent behaviour or specific interactions when you need a stronger basis for action. Finally, do not let the score become the only feedback target: Goodhart's Law warns that a measure turned into a target can lose its value, so combine NPS with qualitative comments and other product metrics.
Turn segments into follow-up
Build a dashboard that tracks NPS over time by subscription status, country, device type, geography, and session count. Review movement after releases and feature launches, then drill into segments where a platform bug, localisation issue, or particular user journey may be responsible. Invite promoters to leave an app-store review or join a referral programme while their experience is fresh. For detractors, read the written feedback, offer a relevant incentive such as a discount or free trial where appropriate, and point them to underused features that address the stated problem. CRM automation can route detractors into personalised help, offers, or feature discovery instead of waiting for them to cancel.
Launch checklist
Use this launch checklist: define the single recommendation question; choose an in-app trigger for active users; use individual 0-10 buttons with no default; add an open-ended why prompt; segment free and paid users; create a time-series dashboard; set promoter and detractor follow-up paths; and review NPS beside conversion, engagement, retention, reviews, and support tickets. The objective is not merely to collect a number, but to convert the reasons behind that number into product and retention improvements.
Frequently asked questions
How is NPS calculated?
Subtract the percentage of detractors (0-6) from the percentage of promoters (9-10). Passives (7-8) do not enter the calculation. The result ranges from -100 to +100.
How should an in-app NPS survey be designed?
Use separate 0-10 buttons instead of a slider or pre-selected value, ask one clear recommendation question, trigger it in-app for active users who have used core features, and follow with an open-ended why question.
How can NPS data be used without becoming a vanity metric?
Segment by subscription status, country, device type, geography and session count, track movement after releases, and pair scores with app-store reviews, support tickets, qualitative comments and behaviour questions.