Customer satisfaction survey questions: the six worth asking
Customer satisfaction survey questions: the six a small business should ask, in order, with wording rules, two copy-ready templates and a one-week plan.
Written by OpenRated Team You want to know how your customers think you are doing, and the usual guide to customer satisfaction survey questions hands you a pile of them, 50 or 60 at a time. A customer who paid you yesterday will answer about six. So the question is which six, in what order, and what you do with the answers on Monday.
This guide is for the owner or manager of a small business: a cleaning firm, an agency, a repair shop, a bakery. It gives you the six questions and the decision each one informs, the wording rules that keep the answers honest, when and how to send the survey, two templates to copy, a plan for the week after, and the line between a private survey and a public review request, which is where the rules bite.
Customer satisfaction survey questions: the six to ask, in order
The best customer satisfaction survey questions are the ones whose answers change what you do next. Each of these six feeds one decision, and the order follows the way a customer thinks right after dealing with you: a general feeling first, the detail while it is fresh, then a judgement about the future, then a couple of easy questions about themselves.
1. Overall satisfaction. "How satisfied are you with the work we did for you?" on a five-point scale from very dissatisfied to very satisfied. Ask it first, because it is the one answer you still want if the customer answers nothing else. It tells you whether the business is healthy this month. Tracked over time, this is your customer satisfaction score (CSAT): the share of answers in the top two boxes.
2. How easy it was. "How easy was it to get what you needed from us?" from very difficult to very easy. This is the customer effort question, and it is the sharpest diagnostic of the six. A customer can be satisfied with the result and still have found you hard to book, hard to reach, or hard to pay. A low answer here points at a process (the phone line, the quote form, the hand-off between staff) before it points at a person.
3. One thing to improve. "What is one thing we could have done better?" as a free-text box. The word "one" is doing the work: it asks for the thing that annoyed them most, in their own words, and that is your priority list. "Any other comments?" gets you nothing or an essay.
4. Likelihood to recommend. "How likely are you to recommend us to a friend or colleague?" on a scale of 0 to 10. This is the net promoter question. It measures something different from satisfaction: whether the customer would put their own name behind you, which is what makes word of mouth happen. What counts as a good number, and how the score is worked out from the 0-to-10 answers, is in what is a good NPS score.
5. Who they are. One segment question, so you can aim the fix. Pick the split that matters for your business: first job or repeat customer; which service they bought; which branch or which member of staff served them. A low ease score from first-time customers says the onboarding is the problem; the same score from regulars says something changed.
6. Permission to reply. "May we contact you about your answers?" with yes or no, and a box for a name or an email address. It makes the survey a conversation. Without it, a customer who writes two angry sentences in question 3 never hears from you again, and you never learn what they meant.
Our rule of thumb, and no more than that: questions 1 to 4 are the fewest worth sending, and past seven the answers tend to get careless. Six is the set the templates below use.
Wording rules: one thing per question, no leading, a balanced scale
A list of sixty questions does not tell you what makes any one of them fair. Three rules cover most of it.
One thing per question. "Was our staff friendly and the service fast?" cannot be answered by a customer who met a friendly person and waited forty minutes. Split it, or pick the one you will act on.
No leading. "How much did you enjoy our excellent service?" tells the customer what to say. Describe the thing, ask for the rating, and leave the adjectives out.
A balanced scale. The scale needs as many negative options as positive ones, a neutral middle, labels on both ends, and the same direction every time (low on the left, high on the right). A five-point scale with four positive labels and one negative is a marketing scale, and the customers who notice will stop trusting the rest of the survey.
| Instead of this | Ask this | Why |
|---|---|---|
| Was our staff friendly and the service fast? | How easy was it to get what you needed from us? | One thing per question; ease covers both |
| How much did you enjoy our excellent service? | How satisfied are you with the work we did for you? | No leading adjective |
| Rate us: Good, Great, Excellent, Outstanding | Very dissatisfied, Dissatisfied, Neutral, Satisfied, Very satisfied | A balanced scale with a middle |
| Any other comments? | What is one thing we could have done better? | Asks for the one that matters |
Two smaller rules. Use the words a customer uses, so "the job" or "your visit" rather than "the engagement". And make every question after the first optional: a customer who hits a required box they cannot answer closes the tab.
If you work in a particular trade (construction, cleaning, logistics, hospitality, real estate, food), the six questions do not change. Change the object of question 2 (how easy it was to book, to get a quote, to track the delivery, to check in) and the split in question 5 (which site, which route, which room type). Patient surveys in healthcare carry privacy rules of their own and are outside this post.
When to send it, and by what channel
Send it right after the interaction, while the customer can still remember the detail. For a service job that means the same day the work finished, or the day the invoice went out. For a shop it means at the counter or within a day of the purchase; for a product the customer has to use for a while (a mattress, a piece of software), send it after they have had a fair chance to use it, and say in the message why you waited.
The channel is the one the customer already used with you. A customer who booked by email gets an email; one who texts you gets a text with a link.
| Channel | Best when | Watch out |
|---|---|---|
| Email with a link | You have the address and the job was done remotely or at their home | Subject line says it is six questions and two minutes |
| Text message with a link | The customer texted or called you to book | One message, no follow-up chasers |
| QR code on the receipt, the invoice or a counter card | A shop, a cafe, a salon | The code must open the survey directly, with no login |
| Printed card at the counter | Customers who will not scan a code | Someone has to type the answers up each week |
| A form on your own website | You have no contact details for walk-in customers | Only works if you tell customers it is there |
Whichever channel you use, the message is three lines: who you are, why you are asking, and that it takes two minutes. Do not put the survey in the same message as a public review request; the next section says why.
Two copy-ready templates
Both are original templates, free to copy and change, with the scale labels spelled out so you can paste them into any survey tool, a form on your site, or a printed card.
Template: a service business (a plumber, a cleaning firm, an agency, a repair shop)
- How satisfied are you with the work we did for you? Very dissatisfied / Dissatisfied / Neutral / Satisfied / Very satisfied
- How easy was it to get what you needed from us, from first contact to the finished job? Very difficult / Difficult / Neither / Easy / Very easy
- What is one thing we could have done better? (free text)
- How likely are you to recommend us to a friend or colleague? 0 (not at all likely) to 10 (extremely likely)
- Was this your first job with us? First job / We have used you before
- May we contact you about your answers? Yes, at this email address: ______ / No, thanks
Template: a shop (a bakery, a boutique, a hardware store, a cafe)
- How satisfied are you with your visit today? Very dissatisfied / Dissatisfied / Neutral / Satisfied / Very satisfied
- How easy was it to find what you came in for? Very difficult / Difficult / Neither / Easy / Very easy
- What is one thing that would make your next visit better? (free text)
- How likely are you to recommend us to a friend? 0 (not at all likely) to 10 (extremely likely)
- How often do you shop with us? First visit / A few times a year / Most months / Most weeks
- May we contact you about your answers? Yes, at this email address: ______ / No, thanks
Question 5 is the one to rewrite for your business. A hairdresser asks which stylist; a courier asks which route; a software company asks which plan. Keep it to one question with four or five options.
What to do with the answers within a week
A survey nobody reads is worse than no survey, because the customers who wrote something are now waiting. Give it a week, in this order.
The first day, read every free-text answer to question 3, all of them, before you look at any score. With a small business's volume the average moves with every reply, so the words are worth more than the number for the first few months.
The second day, reply to everyone who answered yes to question 6, lowest scores first. The reply is short: thank them, say what you understood, and say what you will do or why you cannot. A low score answered within two days is your best chance of keeping that customer.
The third day, sort the question 3 answers into buckets. Aim for three to five, which is as many as you can act on; for most trades they look like waiting time, communication, price surprises, the state the job was left in, one member of staff. Count them. The biggest bucket is the fix.
The fourth day, pick one fix you can finish this month and write down the one question whose score should move when it works. If the fix is "confirm appointments by text", the ease score from first-time customers is the one to watch.
By the seventh day, tell customers what changed, in the next survey's invitation or on the counter card. This is the point of the whole exercise: a customer who sees a change has a reason to answer next time.
Keep the scores in a spreadsheet by month and by the segment from question 5. If you run a help desk, most tools in help desk software can send question 1 automatically when a ticket closes, and the customer success tools in customer success software collect the recommend question inside the product; the wider list of categories is at software categories. For a shop, a free form tool and a QR code do the job.
What a survey will not fix: a product people dislike, a price the market will not pay, or a member of staff you already know about. It will tell you which of those it is, faster than a review will, and in private.
A survey is private, a review request is public
The survey above is yours. The answers go to you, the customer is talking to you, and nobody else sees them. A review is the opposite: the customer is talking to your future customers, in public, on Google or on a review site, and the request has rules.
The rule that catches small businesses is gating: sending the survey first, then asking only the customers who scored you 9 or 10 to post a public review. It feels like common sense, and it is against Google's policy and the FTC's guidance. Google's content policy for business reviews lists, under rating manipulation, that merchants must not "Discourage or prohibit negative reviews, or selectively solicit positive reviews from customers", and allows soliciting reviews only "without offering incentives to do so or attempting to influence the rating or the contents of the review" (Google, Maps User Generated Content Policy, read 2026-10-05). In the United States the Federal Trade Commission's guide for marketers on soliciting reviews says the same in plain words: "Don't ask for reviews only from customers you think will leave positive ones", and "If you offer an incentive for a review, don't condition it, explicitly or implicitly, on the review being positive" (FTC, Soliciting and Paying for Online Reviews, January 2022, read 2026-10-05). This is our reading of the published guidance, general information rather than legal advice; what an incentive may and may not be is in is it legal to pay for reviews.
Your own survey is a different matter, and here the rules are ours, not Google's or the FTC's. A thank-you gift for completing it is fine: the answers are yours and nobody is misled by them. Sending it only to the customers you expect to be happy is pointless rather than prohibited: you learn nothing. Steering the answer with a leading question wastes it in the same way.
So keep the two apart. Send the survey to everyone, read it, reply, fix things. Separately, and to everyone as well, ask for a public review in a message of its own, with a direct link, and let the score fall where it falls. The honest way to tilt the odds is to have replied to the unhappy customer's survey first, so they have no need to use a public review to get your attention. What those public reviews do for a business, from search visibility to the sale itself, is in why reviews are important for a business.
The one thing to do this week
Copy the template that fits your business, paste it into whatever form tool you already have, and send it to every customer you served this week, by the channel they used to reach you. Read the answers on day one and reply on day two. Once the survey is running and you have answered the unhappy ones, open the public side: how to get more reviews walks through asking every customer for a review without breaking the rules above.
