Thought Leadership, Digital PR & Content SyndicationPlaybook3 min readUpdated September 2026

What to Do With an NPS Score Below Zero

A negative or unexpectedly low NPS score tends to produce one of two reactions: panic, or dismissing it as noise from a handful of unhappy people. Neither is useful. The number itself is a starting point for a conversation, not a diagnosis, and treating it as the final word skips the step that actually tells you what to fix.

The real value in a survey like this is almost never the single number. It's in the comments underneath it, the pattern across segments, and what you do with the people who scored you low.

One survey doesn't tell you what's wrong, it tells you where to look

A score on its own tells you that something is off, and roughly how many people feel that way, but it says nothing about what. Two companies with an identical negative score could have completely different underlying problems: one from a confusing onboarding flow, the other from a pricing change nobody explained well.

Treat the score as a trigger to investigate, not a conclusion. If you skip straight to a company-wide response before understanding the actual cause, you risk fixing something that wasn't the real problem while the real one keeps quietly driving churn.

Read the verbatim comments before you look at the score again

The open-text comments attached to a survey are almost always more useful than the number itself, and they're the part most teams skim past fastest. Read every comment from anyone who scored you low, in full, before you look at any aggregate chart or trend line.

You're looking for a pattern that repeats across multiple unrelated respondents, not a single detailed complaint from one person. If several different customers mention the same specific friction point in their own words, that's your priority. A single elaborate complaint from one account, while worth reading, doesn't tell you the same thing about the broader base.

A useful decision rule: if the comments keep repeating one theme, fix that theme first. If they scatter across many unrelated issues, look at segment differences before acting. For example, if new customers cite a confusing setup while long-time customers mention price, treat those as two separate problems with two owners, not one company-wide initiative. Write down the theme, the segment, and who owns the fix, then check the same segment on your next survey. That keeps a negative score from turning into a vague promise to do better, which customers rarely notice.

Segment the score before you react to the overall number

An overall score can hide a lot. Break it down by how long someone has been a customer, which plan or tier they're on, and if you have it, which feature set they use most. A company-wide score that looks mediocre can be masking a segment that's genuinely happy sitting alongside a segment that's actively unhappy, and averaging them together tells you neither story clearly.

Say your overall score sits near zero, but new customers in their first ninety days score noticeably lower than customers who've been with you over a year. That's not a general problem, it's specifically an onboarding problem, and the fix belongs there, not in a company-wide initiative aimed at everyone.

Close the loop with detractors, even if you can't fix everything

Reaching out personally to anyone who scored you low, even just to ask a follow-up question or acknowledge their feedback, matters more than most teams assume. A large share of detractors who never hear back conclude the survey was performative, which damages the relationship further, independent of whatever the original complaint was about.

You don't need to fix every issue a detractor raises to close the loop well. Acknowledging the feedback, explaining what you can and can't address right now, and following up again once something does change is often enough to meaningfully shift how that customer feels, even without a full resolution.

What a healthy follow-up cadence looks like

Running a full survey too often causes fatigue and declining response rates, while running it too rarely means you're reacting to problems months after they started. A quarterly cadence for most small and mid-sized companies strikes a reasonable balance, with a shorter, targeted check-in after any major product or pricing change that could shift sentiment quickly.

Track response rate over time alongside the score itself. A declining response rate, even with a stable score, often means the survey itself is starting to feel routine and ignorable, which is worth addressing with the format or timing before it erodes the quality of the data you're collecting.

After a low score, work through these steps in order:

  1. Read every comment from customers who scored you low, in full, before looking at any aggregate chart or trend line.
  2. Look for a friction point that repeats across several unrelated respondents, not one detailed complaint from a single account.
  3. Segment the score by how long someone has been a customer, their plan or tier, and the features they use most.
  4. Reach out to detractors to acknowledge feedback and explain what you can and can't address right now.
  5. Follow up again once something changes, and track survey response rate alongside the score itself.
Executive Capability Standard

What Good Looks Like

The standard for NPS follow-up is reading every low-score comment for patterns, segmenting the result, and personally closing the loop with detractors, not just reporting the number.

Building The Capability (5-Stage Skill Ladder)

1. Learn:Read every verbatim comment from your last survey's detractors in full and note any theme that repeats across more than one.
2. Do Manually:Personally reach out to your most recent detractors with a short follow-up question about what drove their score.
3. Delegate:Assign a specific owner to close the loop with every detractor within a set number of days after each survey.
4. Automate:Set up segmented reporting by tenure and plan so a mediocre overall score doesn't hide a specific, fixable segment.
5. Buy:Bring in a customer experience specialist if detractor patterns point to a structural onboarding or support problem you can't resource internally.

How to Get Started

Frequently Asked Questions

What's a reasonable NPS score for a small or mid-sized company?

It varies enormously by category, so comparing your score against a broad, blended external norm is less useful than tracking your own trend over time. A score that's improving quarter over quarter matters more than how it compares to a number from an unrelated industry.

Should I survey every customer or just a sample?

For most small and mid-sized companies, surveying everyone is manageable and gives you a complete picture rather than a sample that might miss an important segment. Sampling becomes more useful once your customer base is large enough that survey fatigue from full-population surveys becomes a real concern.

How long should I wait before re-surveying the same customer?

Give at least a full quarter between full surveys for the same customer, longer if nothing significant has changed in their experience. A targeted follow-up after you've addressed specific feedback is different and can happen sooner, since it's checking one thing, not asking the whole survey again.

About the numbers

This guide doesn't quote a sourced benchmark. Figures in it are estimates or general guidance, so check them against your own numbers.

Related Guides