A definition of an NPS detractor: how they’re counted, why they matter more than their number, and how teams win them back. Full disclosure: unitQ publishes this guide and categorizes feedback of the kind described below. 1
A detractor is a survey respondent who answers the Net Promoter Score question, “How likely are you to recommend us?”, with a score of 0 to 6 on the 0-to-10 scale. Detractors subtract directly from NPS, which is calculated as the percentage of promoters (9 to 10) minus the percentage of detractors. Beyond the arithmetic, a detractor is an at-risk customer who has told you, in their own words if the survey asks, exactly what is pushing them away.
How detractors are counted
NPS divides respondents into three bands. Passives are ignored by the formula, so the score moves only when respondents cross into or out of the outer bands.
Promoters
Enthusiasts likely to recommend you and fuel growth.
Passives
Satisfied but unenthusiastic; ignored by the formula.
Detractors
At-risk customers who would not stake their reputation on you.
A company with 40 percent promoters and 25 percent detractors has an NPS of 15. The detractor band is deliberately wide. A 6 might feel like a near-miss, but the methodology treats mild disappointment and open frustration as the same directional signal: this customer would not stake their reputation on recommending you.
Many teams track detractor rate as its own metric rather than watching only the composite score, because the composite can hide offsetting movement. NPS can hold steady while detractors and promoters both grow, which is a very different business than one where both are shrinking.
Why detractors matter more than their count
The number understates the value. Detractor verbatims are disproportionately specific: promoters tend to write “love it,” while detractors write “the app logs me out every time I switch networks, and support took four days.” One of those comments is actionable.
Detractors also cluster. When a batch of low scores cites the same theme, a price change, a redesign, a reliability problem, you are looking at a root cause with a measurable blast radius, not a collection of individual grievances. Categorizing detractor comments by theme, the same way platforms like unitQ categorize all feedback channels, turns a score into a to-do list.
And detractors talk. The premise of NPS is that recommendations drive growth; the mirror image is that detractors actively warn people off. Winning one back removes a negative voice and often creates an unusually loyal advocate, because the recovery itself was the experience.
Winning them back
The pattern that works has three moves:
- Respond quickly and personally. A detractor who hears from a human within a day or two frequently softens, while silence confirms their score.
- Fix the actual cause. An apology without a change just resets the timer on the next low score.
- Close the loop. Tell the customer what changed because of what they said.
For high-value segments, a short follow-up interview digs out context a survey box never captures, and teams increasingly recruit those conversations directly from their detractor lists.
A worked example
A password manager sees its detractor rate climb after a browser-extension rewrite. Reading the verbatims by theme shows most low scores cite autofill failing on a handful of popular sites. The fix is a compatibility patch plus a short personal reply to every detractor who named the bug, linking the release note. Detractor rate falls back within two survey waves, and several of those customers resurvey as passives or promoters. The lesson isn’t the score; it’s that the low scores pointed at one fixable cause once someone read them as a group.
See how unitQ compares on your data
A short demo, run on your own feedback.
Related terms
FAQ
Turn your detractors into a fix list
Look up any app’s free public unitQ scorecard, or take a demo to theme your low-score verbatims and route them to owners.
Sources 1 references
unitQ, "AI categorization of survey verbatims alongside tickets, reviews, and social, turning low-score comments into themed, owned issues." unitq.com. Accessed August 2026.