Form questions
Customer satisfaction survey questions
Ask six to ten, not thirty. Three are never skipped: one overall satisfaction rating on a five point scale, one question about the specific interaction being measured, and one open question asking what to change. Everything else on this page is a spare part you swap in when a particular month needs it.
A satisfaction survey has to find out two things at once: how the customer feels now, and which part of the experience put them there. These thirty questions separate the score from the diagnosis, so the number is comparable month to month and the reasons behind it are still usable.
The questions
Overall
How satisfied are you with your most recent experience with us?Single choiceRequired
The anchor question. Five labelled points from very dissatisfied to very satisfied, and the same wording every month or the trend is worthless.
How satisfied are you with the product itself, separately from the service?Single choice
1 to 5, very dissatisfied to very satisfied.
How satisfied are you with the value for what you paid?Single choice
1 to 5, very dissatisfied to very satisfied.
Has your view of us changed in the last six months?Single choice
Improved a lot, improved, no change, got worse, got much worse.
How does this compare with the last similar company you dealt with?Single choice
Much worse, a little worse, about the same, a little better, much better.
How confident are you that a problem would be sorted if one came up?Single choice
1 to 5, not at all confident to completely confident.
The last interaction
What were you trying to do?Single choice
Buy something, get help with something, change an order, ask a question, something else.
Did you get it done?Single choice
How many times did you have to get in touch before it was finished?Single choice
Which channel did you use?Dropdown
How long did the whole thing take, from start to finish?Dropdown
Was the time it took what you expected?Single choice
Faster than expected, about right, slower, much slower.
Did you have to explain the same thing more than once?Single choice
How clearly was the next step explained to you?Single choice
1 to 5, not at all clearly to completely clearly.
Did anybody follow up without you having to chase?Single choice
How well did the person you dealt with understand what you needed?Single choice
1 to 5, not at all to completely.
If you looked for the answer yourself first, what happened?Single choice
Found it, found something out of date, found nothing, did not look.
What to fix
What is the main reason for the score you gave?Long textRequired
The one question that makes the score mean something. Asked straight after the rating while the reason is still attached to it.
What is the single thing you would change first?Long text
What should stay exactly as it is?Long text
Worth asking. Teams cut the thing customers valued most about twice a year without this question.
Was there a point where you nearly gave up?Long text
Which of these would improve things most for you?Multiple choice
Faster replies, clearer pricing, longer opening hours, better self service, fewer handovers, nothing needs improving.
How easy was it to get this sorted?Single choice
1 to 5, very hard to very easy. Keep it if you are not running a separate effort survey.
Is there anything you expected us to ask about that we have not?Long text
About you
How long have you been a customer?Dropdown
How often do you buy from us?Dropdown
Which product or plan are you on?Dropdown
Are you answering for yourself or for an organisation?Single choice
May we contact you about what you wrote?Single choice
Ask this before the email field rather than after it, so a no means the email is never collected.
Your email, if you said yesShort text
Short text, not the email component, which renders nothing on a published classic form.
Which questions to make required
Only two questions carry the decision, so only two are required: the overall satisfaction rating, because without it the response cannot be counted at all, and the open question that explains the score, because a number with no reason cannot be acted on. Everything else is optional, and a respondent who abandons at a required demographic question costs you both.
- How satisfied are you with your most recent experience with us?
- What is the main reason for the score you gave?
The form built from these questions
A static copy of the finished form, from its own example page.
Three ways to change this bank
For a clinic, replace the last interaction section with the appointment: how long from booking to being seen, whether the reason for the visit was understood, and whether the next step was written down. Keep the overall rating wording identical so the two versions stay comparable. For a short version, cut to five: the overall rating, did you get it done, how easy was it, the main reason for your score, and may we contact you. That fits one phone screen and still answers the only question a monthly report asks. For a follow up wave, send four weeks later to the people who scored one or two, and ask three things: has the issue been resolved, did anybody come back to you, and would you rate it differently now. Do not re-ask the original rating inside the follow up, because the two numbers get averaged together by whoever builds the chart.
Questions people ask
How many questions should a customer satisfaction survey have?
Six to ten, and closer to six when it is triggered by a transaction. The thirty here are a bank to choose from, not a survey. Every question past about ten costs completions, and the ones dropped are at the end, which is exactly where most people put the open question that explains the score.
Should a satisfaction survey be anonymous?
Transactional ones usually are not, because the response is attached to the order or ticket it is about, and that is what makes it useful. Say so plainly on the first screen rather than implying anonymity you cannot deliver. If you want anonymous answers, remove the contact question and the email field, and publish the form as a public link with no login required.
How often should this go out?
Transactional surveys go out once per completed interaction, with a cool down so nobody is asked twice in a month. Relationship surveys, which ask about the company rather than one event, suit six or twelve month intervals. Running both is fine only if the overall rating is worded identically in each.
What response rate should I expect?
Transactional email surveys commonly land between five and fifteen per cent, and in app prompts run higher because they arrive as the task finishes. Treat any published figure as a starting expectation rather than a target. Your own first three sends are the only benchmark worth anything, and the trend beats the level.
Where does the open question go?
Immediately after the rating, while the reason for the number is still in the respondent's head. Put it at the end and you get shorter answers from fewer people, because by then they are finishing rather than explaining. One open question per survey is the working limit; two produce one good answer and one blank.
How do the answers come out when exported?
One column per field in the CSV, with submission time first. Choice questions export the option labels rather than stored values, and a rating grid comes out as one column with each statement and its answer joined together, so plan a parsing step before averaging a grid in a spreadsheet.
Build this form in about a minute
The button opens the form generator with this form already described. Change the wording to match your own, generate, then edit any field.
Create a form with OneCraftOther question banks
Post purchase survey questions
A post purchase survey covers four jobs owned by four different teams, which is why it is written in sections rather than as one list. Split that way, the merchandiser, the checkout team, the logistics lead and the buyer each get a block they can read on their own.
Website feedback questions
A feedback popup gets one screen and about ninety seconds, so the first two questions have to carry the whole survey. Ask the task and whether it was completed, and everything after that is optional detail you can afford to lose.
Loyalty program survey questions
A rewards program is usually reported on sign ups, which is the number that never falls. These questions measure redemption and friction instead, and they make members pick three rewards rather than approve all of them.
On writing the questions themselves, read How to write survey questions. For everything the builder can do, see the form maker. Step by step: Create a conversational form with AI, then View, export and manage your responses.
Written and checked by the OneCraft team. Last checked .