Voice of the Customer Survey: 25 Questions That Get Real Answers (2026)

A good Voice of the Customer survey is short, specific, and writes itself. These 25 questions are organised by what each one is meant to learn — satisfaction, loyalty, effort, churn risk, product direction, or unstructured feedback — and worded to get honest answers, not polite ones. Mix and match what fits the survey you are designing. Skip what does not.

Most VoC surveys fail for one of three reasons: they are too long (so people abandon them), they are too generic (so the answers are useless), or they never ask a follow-up (so the team sees the number without the cause). Every question below is designed to dodge all three traps. Each one is paired with a use case and a "why this wording" note so you can adapt it to your business without losing the intent.

For the broader picture of how surveys fit into a VoC program, see our pillar guide on Voice of the Customer. For deeper coverage of when to use CSAT vs NPS vs CES, see our metric comparison guide.

Before you write a single question

Three rules that make every VoC survey better:

• Keep it short. Five questions is the upper bound for in-the-moment surveys. Three is better. Long surveys get long-form abandonment, not long-form answers.
• Always include one open-text question. The number tells you what changed. The verbatim tells you why. If you only have room for one, make it open-text.
• Tell the customer what you will do with the answer. "Your reply goes to the team that owns this workflow" earns more honest answers than "Help us improve."

CSAT questions (5)

Use these after a single, bounded interaction. A closed support ticket. A purchase. A phone call. Always pair with one open-text follow-up.

1. The standard CSAT

"How satisfied were you with the support you received today? 1 (very unsatisfied) to 5 (very satisfied)."

The default. Works everywhere. Use a 5-point scale unless your tooling forces something else.

2. CSAT for a specific channel

"How satisfied were you with our live chat support? 1 to 5."

Good when you want to compare quality across channels (email vs chat vs phone). Naming the channel inside the question makes the responses cleaner.

3. CSAT for an outcome

"Did we solve your problem today? Yes / Partially / Not yet."

A simpler variant that gets at first-contact resolution without scoring an agent. Useful when you want the answer about the outcome, not the experience.

4. CSAT verbatim follow-up

"What is the one thing we could have done better?"

Better than "any feedback?" because it forces a specific answer. The single best follow-up question on any CSAT survey. Read the responses weekly.

5. CSAT for a self-service flow

"Was this help article useful? Yes / No / Partially. (Optional) What were you trying to do?"

Captures CSAT on knowledge-base content. The optional task description is gold for content gap analysis.

NPS questions (5)

Use these quarterly across the whole customer base, or after a major lifecycle event (90-day-in, post-renewal, post-onboarding). NPS without a follow-up question is half a metric. Always include one.

6. The standard NPS

"On a scale of 0 to 10, how likely are you to recommend [Company] to a friend or colleague?"

The 2003 original. Still the cleanest single number for brand-level loyalty. The 0–10 scale matters — do not switch to a 1–5 scale and call it NPS.

7. NPS verbatim follow-up

"What is the main reason for your score?"

The most important question on any NPS survey. The score gives you the trend. The verbatim gives you the cause.

8. NPS for promoters specifically

"(Shown only to 9–10 scorers) What is the one thing that has made the biggest difference for you?"

Promoter feedback is your marketing copy. Surface it. Quote it. The responses also tell you which features matter most to the people who refer you.

9. NPS for detractors specifically

"(Shown only to 0–6 scorers) What would need to change for you to recommend us?"

Detractor feedback is your roadmap. The wording matters — "what would need to change" feels less accusatory than "why didn't you score higher" and gets longer, more honest replies.

10. NPS for a specific product line

"On a scale of 0 to 10, how likely are you to recommend [Product] to a colleague?"

Useful in multi-product businesses where the overall NPS hides a struggling product. Naming the product inside the question keeps the data clean.

CES questions (5)

Use these right after a workflow you suspect has friction. Sign-up. Account setup. A support resolution. CES is narrower than CSAT and NPS — do not over-apply it.

11. The 2010 Harvard Business Review CES

"The company made it easy for me to handle my issue. (Strongly disagree to strongly agree, 1–7)."

The original CES wording. Still works. The reverse-scored format makes the number harder to read at a glance, which is why many teams swap to the direct variant below.

12. The direct CES

"How easy was it to resolve your issue today? 1 (very difficult) to 7 (very easy)."

The cleaner modern wording. Higher score = easier. Most small teams should use this version.

13. CES for a specific workflow

"How easy was it to sign up for [Product]? 1 (very difficult) to 7 (very easy)."

Naming the workflow in the question gets you data you can act on. "Sign-up" is replaceable: account setup, ticket submission, refund, integration, password reset.

14. CES with the "where did you struggle" follow-up

"Where, if anywhere, did the process feel harder than it should? (Open text)"

The verbatim that turns CES from a metric into a product roadmap input. Read the responses monthly.

15. CES for self-service

"How easy was it to find what you were looking for? 1 to 7."

Captures CES on your knowledge base, help centre, or in-product search. Pair with a list of the top help articles to see which one is causing friction.

Churn risk and retention questions (4)

Use these at risk-of-renewal moments — pre-renewal surveys, cancellation flows, or quarterly NPS surveys. The goal is not the number, it is the verbatim.

16. The renewal intent question

"How likely are you to renew your subscription when it comes up? 1 (very unlikely) to 5 (very likely)."

Best run 60–90 days before renewal. Anything below 4 gets routed to customer success for a check-in.

17. The cancellation reason

"What is the main reason you are cancelling? (Pick one) Price / Missing feature / Switching to a competitor / Project ended / Customer support / Other."

Make the reason required on the cancel flow. The free-text "other" box gets the answers nobody anticipated.

18. The "what would have kept you" question

"(After cancellation reason) Is there anything we could have done that would have changed your mind?"

The most honest moment in any customer journey. People who have already cancelled have no reason to spare your feelings. Listen.

19. The competitor question

"(Optional) Are you switching to a specific competitor? Which one?"

Many teams skip this because it feels uncomfortable. Do not. Knowing which competitors are pulling your customers is one of the most valuable inputs to your product strategy.

Product and roadmap questions (3)

Use these in quarterly or product-focused surveys. Not the primary VoC instrument, but useful for routing roadmap inputs to product teams.

20. The "what's missing" question

"If you could add one thing to [Product], what would it be?"

Forces a single specific request rather than a wishlist. The patterns across 100 responses are more revealing than any single answer.

21. The "what would you remove" question

"If you could remove one feature from [Product], what would it be? Why?"

Almost no team asks this. The answers tell you which features cause confusion or friction. The "why" matters more than the feature itself.

22. The job-to-be-done question

"What were you using before [Product]? What problem were you trying to solve?"

Best run during onboarding. The answers shape positioning, sales pages, and onboarding flows. Read 30 responses in a row and your homepage rewrites itself.

Open-text "tell us anything" questions (3)

Use these as the last question on any survey. Always optional. Never required.

23. The "anything else" sweeper

"Anything else you would like to tell us? (Optional)"

The most-used VoC question. Gets the bonus feedback that did not fit anywhere else.

24. The "what surprised you" question

"What is one thing that surprised you about working with us, good or bad?"

A great unlock for unstructured insight. The "surprise" framing gets answers that "feedback" never would.

25. The forward-looking question

"Where do you want [Product] to be in 12 months that it is not today?"

Best used quarterly with your most engaged customers. The answers shape long-term roadmap conversations and give you direction nobody else has.

How to actually use these questions

Pick three for a CSAT survey. Pick two for an NPS survey. Pick two for a CES survey. Always include one open-text question. Send each survey to a defined moment in the customer journey, not a generic "feedback request" that nobody opens.

The 25 questions above are a menu. The mistake is treating them as a checklist. Use only the ones whose answers your team has the bandwidth to read.

For deeper coverage of which metric to use when, see our VoC vs CSAT vs NPS vs CES guide. For the broader picture of how surveys fit alongside support tickets, online reviews, and behavioural data inside a working VoC program, see our Voice of the Customer pillar.

For the follow-up step that turns survey responses into retention — going back to the specific respondent — see our guide on how to close the customer feedback loop. And when the spreadsheet runs out, our Voice of the Customer tools guide for small support teams covers what to look for in a dedicated platform.

For more focused CSAT and NPS wording examples (50+ across different industries), see our existing guide on the best survey questions to measure customer satisfaction.

Frequently asked questions

How many questions should a Voice of the Customer survey have?

Three to five questions for in-the-moment surveys (post-ticket CSAT, post-workflow CES). Six to eight for quarterly NPS surveys. Anything longer gets abandoned. Response rate drops sharply past five questions in mobile contexts.

What is the best Voice of the Customer survey question?

There is no single best question. The most useful one for your team depends on what decision the survey is going to inform. If forced to pick one, the post-CSAT open-text follow-up ("What is the one thing we could have done better?") gets more actionable insight per minute than any other VoC question.

Should every VoC survey have an open-text question?

Yes. The number tells you what changed. The verbatim tells you why. A survey without open-text feedback is a metric without a diagnostic. Always include at least one open-text field, and read at least 20 responses per month.

How often should you send a VoC survey?

CSAT after every closed support interaction. NPS quarterly. CES after the specific workflow you are diagnosing. Avoid sending the same customer more than one survey per week — survey fatigue is the fastest way to kill response rates.

What is the difference between a VoC survey and a customer feedback survey?

A VoC survey is a subset of customer feedback. VoC specifically refers to surveys that feed into a Voice of the Customer program — meaning the responses are tagged, analysed monthly, routed to the team that can act, and closed back with the customer. A customer feedback survey that nobody reads is not a VoC survey.

Can you use the same questions for B2B and B2C?

Mostly yes. CSAT, NPS, and CES wording transfers across both contexts. The differences are in the lifecycle questions (B2B has longer renewal cycles), the cancellation flow (B2B churn often involves multiple stakeholders), and the timing of NPS (B2B can run quarterly; B2C may need shorter cycles for fast-moving categories).

Ship a survey this week

Pick three questions from this list. Wire them into your most-used customer touchpoint (the post-ticket page, the email after a purchase, the in-product modal after sign-up). Read the first 30 responses with the team. Then iterate.

The biggest gap between teams that run a working VoC program and teams that do not is almost never the question wording. It is the act of reading the responses every week and acting on what you learn. The questions above remove the writing-from-scratch tax. The rest is the work — and the discipline that holds it together is the one-page customer feedback strategy that says who reads what and what action follows each signal.

If your team handles support through email, every closed ticket is already an implicit feedback signal. Adding a one-question CSAT to the close-out email turns the implicit into a measurable. SupportBee's shared inbox makes that wiring trivial and keeps every response tied to the ticket history that produced it. Start a free 14-day trial to set up your first VoC survey.