Home / TinyForms / Customer feedback survey
Customer feedback survey: how to create one people actually answer.
The complete guide: which metric to use (NPS, CSAT, or CES), how to build a survey that gets finished, 45 ready-to-use questions organized by goal, and the response-rate benchmarks to judge yourself against. Plus the part most guides skip, which is what you do with the answers.
Pick one metric for the moment: CSAT after an interaction, CES after a task, NPS for the relationship. Ask 1–3 questions (a score plus an open “why”), trigger it right after the experience, in the channel the customer is already in. Then act on what comes back. Route unhappy customers to a human the same day, and tell people what changed. The survey is the easy part; the loop is the value.
What is a customer feedback survey?
A customer feedback survey is a short, structured set of questions that measures how customers feel about your product or a specific experience, and why. A good one pairs one scored metric (NPS, CSAT, or CES) with one open “why”. That gives you a trackable number plus the reason behind it.
That pairing matters more than anything else on this page. A score without a “why” tells you something changed but not what to fix; open comments without a score can’t be tracked over time. Every feedback program that works, from a two-person startup to an enterprise CX team, runs on the same primitive: number + reason, captured close to the moment.
What a feedback survey is not: market research (different discipline, longer instruments), a lead-capture form wearing a disguise, or an annual 40-question event. Customers can tell the difference, and your completion rate will tell you they can.
NPS vs CSAT vs CES: match the metric to the moment
CSAT measures satisfaction with one interaction; use it after support or delivery. CES measures effort and belongs after onboarding or a task. NPS measures the overall relationship, on a cycle. Not after every touchpoint. Most teams need two of the three, not all of them everywhere.
| Metric | The question | Scale | When to send | What a result means |
|---|---|---|---|---|
| CSAT | “How satisfied were you with [this experience]?” | 1–5 | Minutes after a support ticket, delivery, or call | % who answer 4–5. Tracks the quality of one touchpoint. |
| CES | “How easy was it to [do the thing]?” | 1–7 | Right after onboarding, setup, or a task | Effort predicts loyalty better than delight does: high effort, high churn. |
| NPS | “How likely are you to recommend us?” | 0–10 | Quarterly, or at relationship milestones | % promoters (9–10) minus % detractors (0–6). The relationship trend line. |
One rule keeps programs sane: each metric answers a different question, so don’t average them together. CSAT can be high while NPS falls (support is great, the product story is losing people). That divergence is information, not noise.
How to create a customer feedback survey in six steps
Decide the one decision the survey will inform, pick the metric for that moment, write 1–3 questions plus an open “why”, add branching so each person only sees what’s relevant, trigger it right after the experience, and wire up what happens to the answers before you send it.
“Should we fix onboarding or pricing first?” is a survey. “Let’s see what customers think” is a fishing trip. If no decision changes based on the answers, don’t send it. Survey fatigue is real, and every pointless ask spends trust.
Interaction → CSAT. Task or onboarding → CES. Relationship → NPS. (Table above.) One metric per survey; resist the committee asking to “just add a few more”.
Surveys with 1–3 questions get 83% completion; the median survey is finished in under two minutes. Steal from the question bank below; every question there is single-purpose and neutral.
Branch on the score: a 9–10 gets “what do you love?”, a 0–6 gets “what went wrong?” and an offer of help. Conditional follow-ups are why well-built longer surveys can still hold high response rates. Every question feels relevant.
In-product beats email roughly 2–3× on response rate (≈27% vs 15–25%, and ~36% on mobile). Trigger CSAT when the ticket closes, CES when setup completes, post-purchase a few days after delivery. A survey sent two weeks later measures memory, not experience.
Decide now: where do responses land, who sees a detractor alert, what email follows a low score? On TinyForms the survey, the response table, the routing workflow, and the follow-up email live on one canvas, so the answers arrive somewhere actions can happen.
45 customer feedback survey questions, organized by goal
Don’t ask all 45. Pick one group that matches your moment, take the scored question plus 1–2 follow-ups, and always include one open-ended question. Every question here is neutral, single-purpose, and short enough for mobile.
- On a scale of 0–10, how likely are you to recommend us to a friend or colleague?
- What's the main reason for your score?
- What's the one thing we could do to raise your score by a point?
- Which product or feature do you find most valuable, and why?
- If you stopped using us tomorrow, what would you miss most?
- How satisfied were you with your experience today?
- How satisfied are you with the resolution of your issue?
- Did we resolve your issue on the first contact?
- How would you rate the friendliness and knowledge of the person who helped you?
- How satisfied are you with how long it took to get a response?
- What could we have done to make this experience better?
- Is there anything still unresolved that we should follow up on?
- How easy was it to get the help you needed today?
- How easy was it to get started with the product?
- Do you agree that we made it easy to handle your issue?
- What was the most frustrating or confusing part of the process?
- Where did you get stuck, even briefly?
- How disappointed would you be if you could no longer use this product?
- What problem were you trying to solve when you found us?
- Which features do you use most often?
- Is there a feature you expected that's missing?
- How would you rate the value for money?
- What almost stopped you from signing up?
- What other tools did you consider before choosing us?
- If you could change one thing about the product, what would it be?
- How was your overall purchase experience?
- Did the product match the description and your expectations?
- How satisfied are you with delivery speed and condition?
- How easy was the checkout process?
- Did you find everything you were looking for today?
- How clear were the setup or getting-started instructions?
- How likely are you to purchase from us again?
- What's the main reason you're cancelling?
- Did you switch to another product? Which one?
- What would have convinced you to stay?
- Which part of the product didn't meet your expectations?
- How likely are you to come back in the future?
- Can we follow up with you about your experience?
- What's the one thing you'd improve first?
- What nearly stopped you from completing this today?
- How would you describe us to a friend, in one sentence?
- What's the best thing about the product?
- What do competitors do better than we do?
- What question should we have asked, but didn't?
- Anything else you'd like us to know?
Want these pre-built? The survey templates gallery has ready-to-use NPS, CSAT, and post-purchase surveys you can customize and publish in minutes. And if 45 questions are not enough, the expanded 50-question bank adds product-market-fit and churn sets with usage rules for each.
Response-rate benchmarks, and the five levers that move them
Typical email survey response rates are 15–25%; above 30% is excellent. In-product surveys average ~27% (and ~36% on mobile) because the ask happens in the moment. Length is the silent killer: completion collapses past the two-minute mark.
- Cut to under two minutes, and say so. “2 questions, 30 seconds” in the invite is the cheapest response-rate lift available. The median completed survey takes 114 seconds.
- Ask in the moment, in the channel. The same survey performs radically differently by channel: in-app ≈27%, mobile ≈36%, email 15–25%. Meet the customer where the experience just happened.
- Make question one a single tap. A score, stars, or thumbs. Momentum carries people through the open “why” that follows.
- Send it from a person. A founder’s or account manager’s name beats a no-reply brand sender, especially in B2B.
- Close the loop publicly. “You asked, we built it” in your changelog or email is why your next survey gets answered. Feedback programs compound or decay based on whether customers see action.
The mistakes that ruin your data
Most bad survey data comes from five self-inflicted wounds: leading questions, double-barreled questions, surveying only your happiest moments, asking without acting, and treating one survey as the whole feedback program.
- Leading questions. “How much did you love our new dashboard?” is a request dressed as a question. Neutral phrasing or nothing.
- Double-barreled questions. “Was the product easy to use and good value?” When someone answers 3, which half did they mean? One idea per question.
- Sampling only sunshine. Surveying only customers who completed onboarding (and not the ones who bailed) produces flattering, useless numbers. Survey the drop-offs too; that’s where the money is.
- Required open-text fields. Required essays are how you turn a 30-second survey into an abandoned one. Make the “why” optional; the people with something to say will say it.
- Collecting without closing the loop. The fastest way to train customers to ignore your surveys is to visibly do nothing with the answers.
The part most guides skip: acting on feedback, automatically
A survey that ends in a spreadsheet is a dead end. The working pattern: responses land in a table, detractors page a human the same day, promoters get a review ask, and themes get tagged for the product team. All of that is automatable. None of it should depend on someone remembering to check.
This is where the tool choice actually matters. A standalone survey tool collects answers; then you’re exporting CSVs and wiring middleware to do anything about them. On TinyCommand, the survey is the front door of a system that already exists: the response writes to a table you own, a workflow routes a 0–6 score to Slack while the customer is still warm, and the follow-up email goes out with the context attached. Compare that to a classic survey tool or Google Forms, where collecting is where the product ends.
If you’re choosing a builder, the test is one question: “what happens in the 60 seconds after someone submits?” If the answer is “it sits in a list,” keep looking. See how the options stack up in our Typeform comparison or the Google Forms alternatives roundup.
Frequently asked questions
What is a customer feedback survey? +
A customer feedback survey is a short, structured set of questions that measures how customers feel about your product, service, or a specific experience, and why. A good one pairs one scored metric question (NPS or CSAT, usually) with one open follow-up. You get a number you can track, plus the reason behind it.
How many questions should a customer feedback survey have? +
Fewer than you think. Surveys with 1–3 questions see completion rates above 83%, and the median survey is completed in under two minutes. A good default is one metric question, one open-ended 'why', and at most three follow-ups. Branching keeps it tight, so each person only sees questions meant for them.
Should I use NPS, CSAT, or CES? +
Match the metric to the moment. CSAT ("How satisfied were you?") measures a specific interaction, so use it after support or delivery. CES ("How easy was it?") measures effort and fits right after onboarding or a task. NPS ("How likely are you to recommend us?") measures the whole relationship. Send that one quarterly, not after every touchpoint.
When should I send a customer feedback survey? +
As close to the experience as possible. Trigger CSAT within minutes of a support ticket closing, CES right after onboarding or first value, post-purchase surveys a few days after delivery, and relationship NPS on a recurring cycle (most teams pick 90 days). A survey sent weeks later measures memory, not experience.
What is a good survey response rate? +
For external email surveys, 15–25% is typical and anything above 30% is excellent. In-product surveys do much better (around 27% on average, and 36% on mobile) because the ask happens in the moment. Channel matters more than copy: the same survey in-app can triple its email response rate.
How do I get more people to answer my survey? +
Five levers move response rates most: keep it under two minutes (say so upfront), ask in the moment and in the channel the customer is already in, make the first question one tap, personalize the invite (who it's from matters), and close the loop, because customers who see action from past feedback answer the next one.
How do I create a customer feedback survey for free? +
TinyCommand's free plan includes unlimited forms and responses. Pick a survey template, customize the questions, and share a link or embed it. Paid plans add the automation around it: enriching respondents, routing unhappy ones to your team, and triggering follow-up email.
Keep going
Build a survey that acts on its answers.
Unlimited surveys and responses on the free plan, with the routing, tables, and follow-up built in.