The main types of customer feedback are feature requests, bug reports, usability feedback, survey scores (NPS, CSAT and CES), open-ended survey answers, support conversations, reviews, sales and churn feedback, interviews, and behavioral signals. Each type tells you something different and needs a different response: a bug needs a fix, a feature request needs a vote count, a low score needs a follow-up. This guide defines each type, shows an example, and says how to act on it.
The examples below are illustrative: we wrote them to show what each type looks like, they are not quotes from real customers. For real, attributed examples from Amazon, Slack, Netflix and others, see 15 customer feedback examples.
This post is the hub of our customer feedback series:
- How to collect customer feedback: channels and a step-by-step setup
- Customer feedback analysis: how to turn feedback into decisions
- Product feedback examples: example messages and an example feedback form
Types of customer feedback at a glance
| Type | What it is | Solicited? | Best channel | First action |
|---|---|---|---|---|
| Feature request | A customer asks for something new | Usually not | Feedback board | Merge duplicates, count votes |
| Bug report | Something is broken | No | Support, in-app widget | Reproduce, prioritize by impact |
| Usability feedback | It works but is hard to use | Either | Interviews, session notes | Watch for repeats across users |
| NPS | 0-10 "would you recommend" score | Yes | Email or in-app survey | Follow up with detractors |
| CSAT | Satisfaction with one interaction | Yes | After a ticket or task | Look at the low scores' comments |
| CES | How easy a task was | Yes | Right after the task | Fix the hardest steps first |
| Open-ended survey answer | Free text in a survey | Yes | Surveys, forms | Tag by theme |
| Support conversation | Questions and complaints in tickets or chat | No | Help desk, live chat | Log product asks to the board |
| Review | Public rating on a review site or app store | No | G2, Capterra, app stores | Reply, then log the request |
| Sales and churn feedback | Why a deal was lost or a customer left | Either | CRM notes, cancel form | Tag by reason, review monthly |
Two more labels cut across this table. Solicited feedback is what you ask for (surveys, interviews). Unsolicited feedback arrives on its own (support tickets, reviews, requests on a board). And every type is either qualitative (words that explain why) or quantitative (scores and counts that show how much). You need both: a score tells you something is wrong, the comment tells you what.
1. Feature requests
A feature request is a customer asking for a capability your product does not have yet. It is the most common type of product feedback and the easiest to lose, because it arrives through every channel at once. See the feature request glossary entry for a fuller definition.
Illustrative example: "Can we export the roadmap as a PDF for our quarterly board meeting?"
How to act on it: Put every request in one place, merge duplicates so the vote count is real, and ask what the customer is trying to do (here: share progress with people who do not log in). Then decide, and tell the people who asked. Saying no with a reason is fine; see the art of saying no to feature requests. FeatureOS feature request software handles the collect, merge and notify steps.
2. Bug reports
A bug report says something is broken: an error, wrong data, a page that will not load.
Illustrative example: "When I filter the board by tag, posts from last week disappear. Chrome on macOS."
How to act on it: Reproduce it, record the steps and environment, and rank by how many people it blocks. Keep bugs apart from feature requests (a separate board works) so a crash never competes for votes with a nice-to-have.
3. Usability feedback
Usability feedback means the feature works but people struggle with it. It often shows up indirectly: "where do I find...", repeated how-to tickets, or hesitation in a user interview.
Illustrative example: "I didn't realize I had to click Save before switching tabs. I lost my changes twice."
How to act on it: One comment is an anecdote; three users tripping on the same step is a pattern. Collect these under one theme and fix the flow, not just the help article.
4. Net Promoter Score (NPS)
NPS asks "How likely are you to recommend us to a friend or colleague?" on a 0-10 scale. Promoters score 9-10, passives 7-8, detractors 0-6. NPS is the percentage of promoters minus the percentage of detractors. Our free NPS calculator does the math.
Illustrative example: Score 4, with the comment "Good product, but the mobile app is too slow to use on site visits."
How to act on it: The number is a trend line; the comment is the work. Reply to detractors within days, tag their comments by theme, and track whether the themes shrink after you ship fixes.
5. Customer satisfaction (CSAT)
CSAT asks how satisfied a customer was with one specific thing, usually on a 1-5 scale, right after it happened: a support ticket, onboarding, a purchase. More in our post on CSAT survey forms.
Illustrative example: 2 out of 5 after a support chat, with "Took three replies to get a straight answer."
How to act on it: Read the low scores first. CSAT is narrow by design, so route each low score to the team that owns that touchpoint.
6. Customer effort score (CES)
CES asks how easy it was to get something done, such as "The import was easy to complete" (agree to disagree).
Illustrative example: "Disagree. I had to reformat my CSV three times before it was accepted."
How to act on it: Send it right after the task while the memory is fresh, and fix the steps with the worst scores. Effort is often a better early warning than satisfaction.
7. Open-ended survey answers
These are the free-text answers in any survey: "What is one thing we should improve?" They are qualitative and the richest source of why. For question ideas, see 11 customer feedback survey questions or draft a survey with our survey generator.
Illustrative example: "I wish I could see which customers asked for a feature before I prioritize it."
How to act on it: Tag each answer by theme and count the themes. That turns a pile of text into a ranked list. The analysis guide shows how.
8. Support conversations
Tickets, live chats and emails to support are unsolicited feedback in volume. Most are questions, but many hide a feature request or a usability problem.
Illustrative example: "Is there a way to bulk-archive old posts? Doing it one by one is taking forever."
How to act on it: Give support a one-click way to log the request against the customer's name. In FeatureOS, agents can create or upvote a post from Intercom, Zendesk or Salesforce without leaving the ticket.
9. Online reviews
Reviews on G2, Capterra, Trustpilot or the app stores are public, unsolicited and read by your prospects.
Illustrative example: 3 stars: "Does the job, but reporting is basic compared to what we used before."
How to act on it: Reply publicly and politely, then log any concrete request ("better reporting") on your board so it is counted with everything else.
10. Sales and churn feedback
Lost-deal notes ("they needed SSO") and cancellation reasons ("too expensive for a team our size") are feedback with revenue attached.
Illustrative example: Cancel form: "We moved to a tool that integrates with our CRM."
How to act on it: Ask for a reason in the cancel flow and tag lost deals in your CRM. Review the reasons monthly next to your feature requests: a request that blocks deals deserves more weight than its vote count alone. Also watch your churn rate.
Interviews and behavioral signals
Two more sources do not fit neatly into a form:
- Customer interviews are solicited, qualitative and slow, but they explain the why behind everything else. Five interviews on one theme usually tell you more than five hundred survey scores.
- Behavioral signals (feature usage, drop-off points, search terms in your help center) are feedback customers never typed. They tell you what people do, which is not always what they say.
How to act on every type: close the loop
Whatever the type, the last step is the same: tell the customer what happened. Closing the feedback loop is what makes people send feedback again. In practice:
- Capture each piece of feedback in one system, whatever channel it came from (how to collect customer feedback).
- Sort it by type and theme, merging duplicates (customer feedback analysis).
- Decide what to build, fix or decline, using votes, revenue and effort. The feature prioritization calculator helps.
- Ship and announce it on a public roadmap and in your release notes.
- Notify everyone who asked.
FeatureOS covers those five steps in one place: boards, AI merge suggestions, roadmaps, changelogs and voter notifications. See customer feedback software for the overview or feedback boards for intake. Plans start at $60/month for 5 seats, and you pay only for team seats; see pricing for details and the 14-day free trial.
