Net Promoter Score has a bad reputation among product teams, and the reputation is partly deserved. Most companies collect NPS quarterly, celebrate if the number is above 50, worry if it is below 30, and do absolutely nothing with the data in between. The score becomes a vanity metric: good for board decks, useless for product decisions.
The problem is not NPS itself. The problem is how companies use it. NPS is useful when you do three things that most companies skip: segment the scores, read the verbatims, and close the loop.
Segmenting the scores means breaking NPS down by cohort, plan, feature usage, and acquisition channel. An overall NPS of 45 is meaningless. An NPS of 65 among Pro plan users and 25 among Starter plan users tells you that the Pro experience is strong and the Starter experience is broken. An NPS of 55 among users who use the daily queue and 20 among users who do not tells you that the queue is your magic feature and non-queue users are not finding value. Segmentation transforms a single number into a diagnostic tool.
Reading the verbatims means actually reading every written response, not just the score. The score tells you how many people are happy or unhappy. The verbatim tells you why. "The creative does not match my brand" is a different problem than "the creative is too slow" is a different problem than "the pricing is too high." Each problem has a different solution. If you only look at the score, you know you have unhappy users. If you read the verbatims, you know what to fix.
Closing the loop means responding to every detractor (score 0-6) within 48 hours. Not with a template. With a specific response that acknowledges their concern and explains what you are doing about it. This is where most companies fail because it requires time and humility. But the ROI of closed-loop NPS is remarkable: 30-50% of detractors who receive a personal response within 48 hours revise their score upward in subsequent surveys. The act of listening converts detractors into neutral users and sometimes into promoters.
I run NPS monthly for mani, which is more frequent than most companies. Monthly NPS captures the trajectory. Quarterly NPS captures a snapshot. The trajectory is more useful because it tells you whether changes you made are working. If you shipped a feature in March and NPS drops in April, the feature might have caused problems. If NPS rises in April, the feature might be driving satisfaction. Monthly cadence creates a causal link between product changes and satisfaction changes that quarterly cadence cannot provide.
The segmentation we find most useful is by feature engagement. Users who engage with the daily queue have an NPS of 62. Users who generate but do not use the queue have an NPS of 34. Users who set up Brand DNA but rarely generate have an NPS of 18. This tells us clearly: the daily queue is the value driver. Users who do not engage with the queue are not finding value. Our activation efforts should focus on getting users into the queue habit, not on adding more features.
The growth lever dimension of NPS comes from promoters (score 9-10). These are users who are willing to recommend you. Most companies do nothing with promoters beyond feeling good about the score. We do something specific: we ask them to share. A follow-up email to promoters that says "You gave us a 9. Would you share your experience on Product Hunt / Twitter / LinkedIn?" converts at 15-20%. That is 15-20% of your happiest users actively promoting you, driven by a single email. No referral bonus. No incentive. Just asking people who are already enthusiastic to share their enthusiasm.
The detractor verbatims are equally actionable for growth. Detractor feedback tells you what to fix to convert unhappy users into happy users. Each fix improves the product for all users, not just the detractor. A detractor who says "the creative is too samey" is voicing a concern that 20 other users have but did not express. Fixing it improves satisfaction for all 20. The detractor's voice is amplified feedback for the silent majority.
Our NPS process is: survey monthly, segment by plan and feature engagement, read all verbatims, respond to all detractors within 48 hours, ask promoters to share, and track the trajectory over time. Total monthly effort: about 3 hours. Impact: a diagnostic tool that tells us exactly what to fix and exactly who to activate. That is NPS as a growth lever, not a vanity metric.
Mani's NPS is currently 48. That is good but not great. The detractors cite creative variety (too repetitive over time) and the promoters cite the daily queue (saves them hours per week). We know what to fix and what to double down on. That is the value of NPS when you use it correctly.
The cultural dimension of NPS practice matters. If the product team sees NPS as a threat (low score means we are failing), they will resist the measurement. If they see it as a compass (the score tells us exactly where to improve), they will embrace it. The framing determines the adoption. At mani, NPS results are shared in the weekly product meeting with the framing: here is what our users told us to fix this month. The tone is constructive, not defensive.