Business Growth
How to Create a Customer Survey (Questions, Template, and Examples)
How to create a customer survey for a small business: how to pick one goal, which question types to use, good and bad survey questions, a short fill-in template, how Net Promoter Score works, when to send it, how to read the answers, mistakes to avoid, and a Claude workflow for drafting and analyzing your survey.

Founder, AI Tools and Training Club · November 1, 2026 · 8 min read
The short version
- A good customer survey starts with one decision you need to make. Write that decision down first, then only ask questions that help you make it.
- Keep it short, mix a few rating questions with one or two open questions, and word every question neutrally so you are not steering people to the answer you want.
- The survey is only worth sending if you act on it. Read every open answer yourself, group them into themes, change one thing, and tell customers what you changed.
How do you create a customer survey?
You create a customer survey by deciding the one thing you want to learn, choosing who to ask, writing a small number of clear, neutral questions, testing the survey on a couple of people, and sending it at a moment when the experience is fresh in the customer's mind. Then you read the answers, group them into themes, and make a change based on what you found.
My strong opinion: most small business surveys fail before they are sent, because they try to learn everything at once. A survey that asks 25 questions about everything gets rushed, half-finished answers. A survey that asks five questions about one decision gets answers you can actually use.
Step 1: Pick one goal for the survey
Before you write a single question, finish this sentence: "After this survey, I will decide whether to..." If you cannot finish it, you are not ready to send a survey yet.
| Survey goal | Decision it helps you make | Who to ask |
|---|---|---|
| Find out why people buy | What to say in your marketing and on your sales page | Recent customers |
| Find out why people leave | What to fix in your product, service, or onboarding | Customers who cancelled or did not renew |
| Check satisfaction after delivery | Whether a process is working or needs changing | Customers right after a project, order, or support case |
| Test a new offer idea | Whether to build it, and at what level | Your best existing customers |
| Measure loyalty over time | Whether things are getting better or worse | All active customers, on a regular schedule |
Survey goals and the decision each one supports
Notice that each goal points to a different group of people. Asking the wrong group is one of the quietest ways to get misleading results. If you want to know why people leave, your happiest customers cannot tell you.
Step 2: Choose the right question types
- Rating questions. "How satisfied were you with...?" on a fixed scale. Quick to answer and easy to compare over time.
- Multiple choice. Good for facts, such as how someone found you or which product they bought. Always include an "Other" option.
- Yes or no. Use sparingly, for simple facts only. They hide the detail you usually need.
- Open questions. "What almost stopped you from buying?" These take longer to answer but often give you the most useful insight, in the customer's own words.
A solid default for a small business is a handful of rating or multiple choice questions, plus one or two open questions. The rating questions tell you what is happening. The open questions tell you why.
Good and bad survey questions
Most survey problems come from the wording. Here are the common mistakes and how to fix them.
| Problem | Bad question | Better question |
|---|---|---|
| Leading | How much did you love our onboarding? | How would you rate our onboarding? |
| Two questions in one | Was our team fast and friendly? | Split into one question about speed and one about friendliness |
| Vague | Any feedback? | What is one thing we could do to make this better for you? |
| Hypothetical | Would you buy a premium version? | What did you have to do yourself that you wish we had done for you? |
| Jargon | How was your experience with our fulfillment SLA? | Did your order arrive when you expected it? |
Bad survey questions and better versions
Customer survey template
Here is a short template for a post-purchase survey. Swap in your own wording, and cut anything that does not serve your goal.
- How did you first hear about us? Multiple choice with an "Other" box.
- What problem were you trying to solve when you bought? Open answer.
- What almost stopped you from buying? Open answer.
- How satisfied are you with what you got so far? Rating scale.
- How likely are you to recommend us to a friend or colleague? 0 to 10 scale.
- What is one thing we could do better? Open answer.
- Can we follow up with you about your answers? Yes or no, with an email field if yes.
Question 2 and question 3 are gold for your marketing. The exact words customers use to describe their problem and their hesitation are the words that belong on your sales page and in your ads. If you have not written your ideal customer down yet, pair this survey with How to Create a Buyer Persona.
How Net Promoter Score works
Question 5 in the template is the Net Promoter Score question. Customers answer "How likely are you to recommend us to a friend or colleague?" on a scale from 0 to 10, and you sort the answers into three groups.
| Score | Group | What it usually means |
|---|---|---|
| 9 or 10 | Promoters | Enthusiastic customers who are likely to stay and recommend you |
| 7 or 8 | Passives | Satisfied for now, but not excited enough to recommend you |
| 0 to 6 | Detractors | Unhappy customers who may leave or talk you down |
Net Promoter Score groups
Your score is the percentage of promoters minus the percentage of detractors. Passives count toward the total number of answers but are not added or subtracted. For a small business, the score itself matters less than the trend over time and the reasons behind it, so always follow the number with an open question asking why they chose it.
When and how to send your survey
- Send it while the experience is fresh. Right after delivery, a support case, or onboarding, not months later.
- Tell people how long it takes. An honest "about two minutes" gets more people to start.
- Explain what you will do with it. People answer more carefully when they know someone will read it.
- Ask in the channel they already use. Email for most businesses, a post or direct message for community businesses.
- Send one reminder, not five. A single follow-up is polite. More than that starts to annoy.
Before sending to everyone, have two or three people take it while you watch or listen. Any question they hesitate on or read twice needs rewriting.
How to read and act on survey results
- Read every open answer yourself. Summaries miss the specific phrases that matter.
- Group answers into themes. Put similar comments together and count how often each theme appears.
- Look for patterns, not single loud answers. One angry reply is a person to follow up with, not a reason to change everything.
- Pick one change. Choose the theme that comes up most and affects buying or staying the most.
- Close the loop. Tell customers what you heard and what you changed. It makes the next survey easier to fill.
Survey results fit naturally into a journey map. If answers keep pointing to the same stage, How to Create a Customer Journey Map shows how to turn that into a fix. Happy customers who give you a 9 or 10 are also the right people to ask for a review, which How to Get Testimonials for Your Online Business walks through.
Customer survey mistakes to avoid
- Asking questions you will never act on.
- Making every question required, so people quit halfway.
- Wording questions to get the answer you hope for.
- Only surveying happy customers, then wondering why the results look perfect.
- Treating a small number of answers as proof. A handful of replies is a clue to explore, not a final verdict.
- Collecting answers and never telling customers what changed.
How to create a customer survey with Claude
Claude is useful at both ends of a survey: drafting neutral questions before you send it, and sorting open answers into themes after. Open the Claude desktop app and work through it step by step.
- Tell Claude the one decision the survey should help you make, who you are asking, and what you sell.
- Ask it to draft a short survey of no more than seven questions, and to flag any question that is leading, vague, or asks two things at once.
- Read the draft out loud and cut anything you would not act on.
- After the answers come in, paste them in and ask Claude to group the open answers into themes, count each theme, and quote a few real answers for each.
- Read the raw answers yourself as a check, then decide on the one change to make.
For a deeper look at the analysis side, see How to Analyze Customer Feedback With AI.
Frequently asked questions
How many questions should a customer survey have?
As few as you need to make one decision. For most small businesses that means a handful of questions, mostly quick rating or multiple choice questions, plus one or two open questions. Every extra question makes people more likely to quit or rush.
What are good questions to ask in a customer survey?
Good questions are neutral, specific, and ask one thing at a time. Strong examples include "What problem were you trying to solve when you bought?", "What almost stopped you from buying?", and "What is one thing we could do better?"
How is Net Promoter Score calculated?
Customers rate how likely they are to recommend you on a scale from 0 to 10. Those who answer 9 or 10 are promoters, 7 or 8 are passives, and 0 to 6 are detractors. The score is the percentage of promoters minus the percentage of detractors.
When is the best time to send a customer survey?
Send it soon after the experience you are asking about, such as right after delivery, onboarding, or a support case, while the details are still fresh. Regular loyalty surveys can go out on a fixed schedule instead.
Can AI help me create a customer survey?
Yes. AI can draft neutral questions, point out leading or confusing wording, and group open answers into themes after the survey closes. You still need to choose the goal, decide what to act on, and read the raw answers yourself.