To create a useful customer feedback survey, choose one customer experience, ask a few neutral questions, publish a form customers can actually open, and assign someone to act on the answers. Start with one rating, one diagnostic question and one optional comment. A small, repeatable survey usually tells a small business more than an ambitious questionnaire nobody finishes.
This guide gives you an original post-purchase question set, a Google Forms setup workflow and a weekly action log. It suits shops and service businesses that want to improve delivery, booking, communication or support. Keep sales prospecting separate: someone who has only downloaded a guide cannot evaluate your completed service.
Contents: Choose a decision · Copy the question set · Build the form · Invite customers · Review responses · Worked example · Launch checklist · FAQs
1. Choose the decision your survey should support
Write down the experience, the audience and the decision before opening a form builder. For example: “Ask customers who received an order this week whether delivery updates were clear, so we can decide what to change in our dispatch messages.” This is specific enough to guide both the questions and the follow-up.
Avoid combining every business problem into one survey. A question about your logo does not help diagnose late deliveries. Product satisfaction, booking convenience and support quality can each need a separate survey at a different point in the customer journey.
| Experience | Who to ask | Decision to make |
|---|---|---|
| Order delivery | Customers after delivery | Improve dispatch messages or delivery expectations |
| Completed appointment | Customers after the appointment | Change booking instructions or preparation information |
| Resolved support request | Customers after resolution | Improve explanations or follow-up |
| New customer onboarding | Customers after their first completed milestone | Clarify what happens next |
Assign an owner and a review date. Decide which problems need a direct response and which belong in the weekly review. A serious unresolved complaint should go to the support process immediately; it should not wait for a monthly spreadsheet meeting.
2. Use a short customer feedback survey question set
The following example is designed for a completed order. Adapt the experience named in each question for your business. The wording and response options are a starting template, rather than a standardized research instrument.
| Question | Suggested response format | Why include it |
|---|---|---|
| Overall, how satisfied were you with your recent order experience? | Very dissatisfied / Dissatisfied / Neither satisfied nor dissatisfied / Satisfied / Very satisfied | Track one consistent overall measure |
| How clear were the delivery updates you received? | Very unclear / Unclear / Neither clear nor unclear / Clear / Very clear / I did not receive updates | Find a specific communication problem |
| Which part of the experience most needs improvement? | Product information / Ordering / Delivery / Support / Nothing needs improvement / Other / Not applicable | Direct the next investigation |
| What would you like us to change? | Optional paragraph response | Let customers explain an issue you did not anticipate |
| May we contact you about this feedback? | Yes / No; optional contact field with an explanation | Separate feedback collection from permission to follow up |
For a service appointment, replace the delivery question with “How clear were the instructions before your appointment?” Keep the same scale and add “I did not receive instructions.” Do not ask respondents to score an experience they did not have.
Ask one thing per question. “How satisfied were you with our speed and friendliness?” creates an ambiguous answer when the service was friendly but slow. Split it only if you will act on both answers; otherwise choose the issue you need to understand first.
Use balanced options and neutral wording. “How much did you love your purchase?” encourages praise. A satisfaction scale allows a negative response. Include an appropriate opt-out such as “Not applicable” where needed, and keep open comments optional. For general question-design principles, see SurveyMonkey’s survey-writing guidance.
3. Build and check the survey in Google Forms
Use an account your business controls, then create a new form. Give it a plain title such as “Feedback on your recent order.” Add a short description explaining the purpose, who will read the answers and whether a reply is available. Avoid claiming the form is anonymous if you collect contact details or connect responses to order records.
Add your questions and select a suitable type for each: multiple choice for one answer and paragraph for an optional comment. Make only essential questions required. Google’s form editing instructions explain how to add questions and change their types.
Review responder access and email collection deliberately. Google Forms’ “Limit to 1 response” setting requires a Google sign-in. That may create friction for an external customer audience. If you leave it off, remember that duplicate responses are possible. Choose the trade-off for this survey, rather than assuming every setting improves data quality.
Leave the respondent results summary off when answers may contain private comments. Keep access to the form and response data limited to the people who need it. Tell customers not to enter passwords, payment details or sensitive information in the comments.
Publish the form and check who can respond, then copy the responder link. Publishing a form makes it accessible to the audience you select; creating it alone is not the same as publishing it. Consult Google’s current publish and share instructions if the controls differ in your account.
- Open the responder link in a signed-out browser or private window and confirm the intended audience can access it.
- Complete the survey on a phone, including one negative response and one skipped optional field.
- Verify that the submitted answers arrive in the correct form.
- Check that comments and other respondents’ answers are not exposed through a results link.
- Label test submissions in your working analysis so they do not influence customer figures.
Connect a response spreadsheet if you want to review answers with the action log below. In the form’s Responses area, use Link to Sheets or the response-destination controls. Check the sharing permissions of both files: removing a form collaborator does not automatically remove their spreadsheet access. Google documents the process in Choose where to save form responses.
4. Invite customers at a consistent point
Send the invitation after the experience you want to measure. A delivery survey sent before the package arrives cannot measure delivery satisfaction. Choose a consistent rule, such as the day after confirmed delivery, and document it so different staff members do not invite customers at arbitrary times.
Use a clear invitation: “We would value your feedback on your recent order. Please use this short survey to tell us what worked and what we should improve. If you need help with an unresolved issue, reply to our support message.” Estimate completion time only after someone has timed a trial of your actual form.
Invite the eligible customer group consistently, including customers who complained. Asking only happy customers makes the answers less useful. Record how many invitations were successfully delivered and where they came from. A public social-media link reaches a different audience from a post-purchase invitation; analyze those sources separately.
For a shop counter, put a QR code beside a short explanation of the survey. Keep a readable link available too. For email or messaging, use your permitted customer communication process and avoid repeated reminders that feel like pressure. Make participation optional.
A feedback request is different from a public-review request. Do not make support conditional on a favorable rating. If you later want to publish a customer comment as a testimonial, get separate permission rather than treating a private form response as permission to advertise it.
5. Review answers and assign the next action
Read the comments before reducing everything to one average. Tag responses by a small set of themes such as unclear timing, confusing product information, damaged packaging or unresolved support. Let one response have more than one theme when appropriate.
Track the response count beside any percentage. If you invite 120 customers and receive 24 usable responses, the response rate is 24 ÷ 120 × 100 = 20%. If 18 of those 24 select “Satisfied” or “Very satisfied,” the positive satisfaction share is 75%. These are illustrative calculations, not measured business results.
Exclude “Not applicable,” missing answers and test submissions from the denominator for the question being scored. Show how many valid answers remain. Do not treat a skipped diagnostic question as a dissatisfied response. Keep a record of the calculation so another person can reproduce it.
A voluntary survey represents the people who answered, not necessarily every customer. Small changes between two small groups may reflect who responded. Use feedback alongside actual support records, delivery exceptions and operational checks before making a major decision.
| Theme or issue | Evidence to check | Owner | Next action | Review date |
|---|---|---|---|---|
| Unclear arrival date | Comments plus dispatch message content | Customer support lead | Rewrite the dispatch message with a realistic delivery window | Next weekly review |
| Confusing product compatibility | Question tags plus product-page wording | Product content owner | Add a clear compatibility note to the relevant page | After the next survey batch |
| Unresolved complaint | Customer permission plus support ticket | Assigned support agent | Contact the customer and investigate the ticket | According to the support process |
Choose one or two changes you can complete, rather than creating a long list with no owner. Keep the original survey wording when comparing batches. If you change a question or scale, record the date and treat the new answers as a new series.
6. Worked example: turn a delivery complaint into a testable change
Imagine a small homeware shop gets several comments saying customers did not know when an order would arrive. The team first checks whether dispatch messages were delivered and whether those messages contained useful timing information. It does not assume that the courier caused every complaint.
The shop finds that its message says only “Your order has shipped.” It updates the message to include the carrier’s tracking link, the delivery window it can responsibly state and a contact route for a delayed parcel. The support owner records the change date and checks a real outgoing message for accuracy.
During the next survey batch, the shop retains the delivery-clarity question and compares the counts of clear, unclear and no-update answers. It also checks related support tickets. An improvement would be a useful signal, but it would not prove that the message alone caused the change; order mix, delivery conditions and response selection may differ.
This closes the loop: collect a specific signal, investigate the cause, change one process and check again. The survey has done its job when it informs an action customers can notice.
Customer feedback survey launch checklist
- One experience, eligible audience and business decision are written down.
- Each question asks one thing and allows an honest negative answer.
- Optional comments and follow-up contact fields are clearly identified.
- Responder access works for customers outside the business.
- The form is published and the responder link has been tested.
- Response summaries stay private where comments could identify customers.
- An owner knows where answers arrive and when to review them.
- An action log records the evidence, owner, change and review date.
When feedback identifies confusing promises in your ads or website, check those promises against the actual customer experience. Seolign’s landing page checklist can help review the offer, page information and contact route. Keep this survey focused on the completed experience; it serves a different purpose from campaign attribution or a lead magnet.
Frequently asked questions
How many questions should a small business customer survey have?
Start with only the questions needed for one decision. The five-item example above is a useful starting point, not a universal rule. Remove a question if you cannot explain what you will do with the answer.
Can I collect feedback without requiring an email address?
Yes, if your access and collection settings allow it. Keep contact details optional when they are only needed for follow-up. Check whether other fields, identifiers or invitation methods could still connect an answer to a person before describing the survey as anonymous.
What is a good response rate?
There is no single target that fits every channel or business. Record your audience, delivered invitations and usable responses, then compare similar batches. A higher response rate does not eliminate selection bias or guarantee reliable conclusions.
Should I offer a discount for completing the survey?
Only if the incentive suits your process and you can explain its terms clearly. Never require a positive answer. An incentive can change who participates, so record it when comparing responses with a batch that had no incentive.
How often should I send the survey?
Use an event-based rule suited to the experience, such as after a completed order or appointment. Limit repeated invitations to frequent customers, and review the process if messages become excessive. A regular weekly review can help a small team turn responses into actions.
Featured photo: Dylan Gillis on Unsplash.