Back to all articlesCustomer feedback

Write feedback questions that distinguish an experience from a suggested solution

Alactic Editorial10 min read

Prepared with AI assistance by Alactic Editorial. Examples identified as hypothetical are illustrative, not customer results.

“How was your experience?” often produces an answer that is pleasant but difficult to use. “Would reminders improve your experience?” produces something more specific, but it puts the business’s preferred answer into the customer’s mouth. Neither tells you reliably what happened.

To write useful customer feedback questions, ask about a specific occasion in this order: the customer’s goal, what they experienced, the consequence, and only then a suggested improvement. Keep those answers separate. A customer can describe a genuine difficulty and propose a solution that would not resolve it; both contributions are useful, but they are different kinds of evidence.

The worksheet below shows how to revise common feedback questions into a short, usable sequence. The aim is not to collect more text. It is to collect answers that let your team distinguish a reported event from an interpretation or an idea.

1. Decide which experience the questions should recover

Choose one bounded experience before drafting questions: a recent booking, a completed support request, or an attempt to find a document. A question about the entire customer relationship invites respondents to combine different occasions, teams and expectations into one judgement.

Use a neutral anchor such as “Thinking about your most recent booking attempt…” rather than “Thinking about the problems with your booking…” The first identifies the event. The second assumes there was a problem.

Also decide who can answer. Someone who abandoned an attempt may have valuable feedback, but wording that assumes completion can exclude them. “Booking attempt” accommodates that experience; “your successful booking” does not.

Before writing, complete this internal sentence: We want to understand what happened when customers tried to [specific task], including attempts that did not finish. If the sentence contains several unrelated tasks, narrow the scope or use separate sequences.

This decision makes answers easier to interpret. It does not make the respondents representative of every customer, and a polished question cannot correct an invitation list that leaves out people whose attempts failed.

2. Give each question one evidence job

A short sequence works when each question collects something distinct:

  • Goal: What was the person trying to accomplish? This establishes their intended outcome, not your preferred workflow.
  • Experience: What do they recall happening? This collects a reported sequence, action or point of difficulty.
  • Consequence: What difference did that experience make? This separates inconvenience from an effect on the task.
  • Suggestion: What would they change? This captures an idea, preference or proposed remedy.

Avoid combining these jobs in “How easy was booking, and what features should we add?” A respondent might answer only the second half. You would then have an idea without knowing which experience prompted it.

Reported experience is still self-report, not independently verified observation. “I never received confirmation” records what the customer recalls receiving or noticing. It does not, by itself, establish whether a message was sent, received, overlooked or filtered. Preserve that distinction when summarising feedback.

Likewise, “add a reminder” establishes that someone suggested a reminder. It does not establish that reminders are missing, that other customers need them, or that a reminder would solve the underlying difficulty.

3. Filled hypothetical question-revision worksheet

Fictional worked example: A service business wants feedback on appointment booking. Its team suspects that extra messages and a new booking app would help. The following worksheet is hypothetical; it contains no customer results. The revised wording deliberately withholds those ideas until the customer has described their own experience.

Worksheet entry: replace broad satisfaction with a goal

Original question: “How satisfied were you with our excellent booking service?”

Problem: The praise cues a favourable answer. “Booking service” is broad, and a satisfaction judgement does not explain what the person needed.

Improved wording: “Thinking about your most recent appointment booking attempt, what were you trying to arrange?”

Optional probe: “Was there anything the appointment needed to fit around?”

Evidence it can collect: The customer’s intended outcome and relevant constraints, such as arranging a particular service around their availability.

What it cannot establish: Whether the booking process worked or which change would improve it.

Worksheet entry: replace a presumed cause with an account

Original question: “How confusing was the booking page?”

Problem: It assumes both confusion and use of a particular page. A customer who booked another way cannot answer accurately.

Improved wording: “What happened as you tried to arrange that appointment?”

Optional probe: “Was there a point where you were unsure what to do next, or did it feel clear throughout?”

Evidence it can collect: A recalled account of the attempt, including the route used, actions taken and any uncertainty.

What it cannot establish: A technical cause, or whether the same event happens to other customers.

Worksheet entry: separate the consequence from the complaint

Original question: “How frustrating and time-consuming was the process?”

Problem: It bundles two judgements and assumes both were negative. Frustration and time spent may differ.

Improved wording: “What difference, if any, did that experience make to getting the appointment arranged?”

Optional probe: “What did you do next?”

Evidence it can collect: A reported consequence, such as continuing without difficulty, checking elsewhere, contacting the business or stopping the attempt.

What it cannot establish: An exact amount of lost time or the business-wide cost of the experience.

Worksheet entry: open the improvement question

Original question: “Would a new app and helpful reminders make booking easier?”

Problem: It presents two solutions together and describes one positively. Agreement cannot tell you which idea the respondent supports or why.

Improved wording: “What, if anything, would you change about arranging the appointment?”

Optional probe: “Which part of your experience would that change address?”

Evidence it can collect: A customer-generated suggestion and its stated connection to the experience.

What it cannot establish: That the proposed change is necessary, feasible or effective.

Worksheet entry: repair an ambiguous reference

Original question: “Was it quick enough?”

Problem: “It” could mean finding an appointment, completing the booking or waiting for confirmation. “Enough” has no explicit reference.

Improved wording: “After you submitted the booking request, what happened before you knew whether the appointment was confirmed?”

Optional probe: “What were you expecting to happen at that point?”

Evidence it can collect: The reported confirmation experience and the expectation against which it was judged.

What it cannot establish: Actual elapsed time unless the customer supplies a basis for it.

Use condition: Ask this only when the person says they submitted a request. Otherwise, it assumes a step they may never have reached.

4. Assemble the shortest sequence that still explains the event

For the hypothetical booking example, the core sequence can be copied into a feedback form or interviewer’s guide:

  1. Thinking about your most recent appointment booking attempt, what were you trying to arrange?
  2. What happened as you tried to arrange that appointment?
  3. What difference, if any, did that experience make to getting the appointment arranged?
  4. What, if anything, would you change about arranging the appointment?

Introduce it with a brief scope statement: “Please think about your most recent attempt, whether or not you completed the booking. We want to understand what happened; suggestions are welcome but not required.”

These are open questions, so they require more effort than selecting a rating. Do not automatically add every worksheet probe as another required field. Use probes only when an answer leaves an important gap. In a self-completed form, test whether a short hint helps before adding more questions.

Allow respondents to say they do not remember, that nothing needs changing, or that a question does not apply. Requiring an improvement can manufacture criticism just as leading praise can manufacture approval.

If the business also needs a satisfaction rating, keep it separate from the evidence sequence. Do not treat a score as a substitute for an account of what happened.

5. Probe without supplying the missing answer

A useful probe asks for detail within the customer’s account. A leading probe supplies a cause, consequence or solution they have not mentioned.

In a hypothetical response, a customer says, “The confirmation was unclear.” A neutral follow-up is “What left you unsure?” A leading follow-up is “Was that because we did not send enough messages?” The latter moves straight from uncertainty to the team’s messaging theory.

Useful probes include:

  • “What happened just before that?”
  • “What did you see or receive at that point?”
  • “What did you do next?”
  • “When you say ‘slow’, which part do you mean?”
  • “What would that suggested change help you do?”

If the customer immediately proposes a solution, acknowledge it without endorsing it: “You suggested a booking app. What happened during your recent attempt that led you to that idea?” Then return to the sequence.

Do not turn clarification into a demand that the customer defend their view. If they cannot recall an event, record the suggestion as a preference without a supporting account. That is more accurate than repeatedly prompting until a plausible story appears.

6. Pretest interpretation, not just completion

A small pretest checks whether people understand the questions as intended. It does not establish how common an experience is across your customer base.

Hypothetical pretest plan: Ask four people with a relevant recent experience to complete the sequence. Include different outcomes where practical, such as a completed booking and an abandoned attempt. After they answer, ask each person to explain what they thought each question meant.

Check these observable points:

  • Event anchor: Did they describe one attempt or combine several?
  • Meaning: Did “arrange an appointment” refer to the same task you intended?
  • Answer separation: Did the experience answer describe events, or only give a general judgement?
  • Assumptions: Did any question require a step they had not taken?
  • Burden: Where did they hesitate, skip an answer or repeat themselves?

Keep a revision log containing the wording, the misunderstanding and the proposed change. If respondents mix several occasions, strengthen the event anchor. If they repeat the same answer for experience and consequence, check whether the consequence question is clear enough—or whether the experience genuinely had no further effect.

Try changed wording with another small group rather than assuming the edit worked. Record observed completion time instead of advertising an untested duration.

7. Preserve the evidence boundary when reviewing answers

Question design loses its value if every answer is later reduced to a feature request. Keep a simple review record with separate fields for goal, reported experience, reported consequence, suggestion and uncertainty.

Hypothetical filled record: A respondent wanted to secure an appointment before rearranging work. They reported submitting a request, remaining unsure whether it was accepted and contacting the business to check. They suggested an app. The uncertainty is what confirmation information they received or noticed.

The defensible summary is: “This respondent reported uncertainty after submitting a request and made an additional contact to check.” It is not: “Customers require an app.” The suggestion may be worth exploring, but the account does not verify it as a requirement.

Keep summaries traceable to the original response. Where the event and suggestion do not connect, mark that gap rather than filling it with the team’s preferred explanation. This is the handoff boundary for feedback design; choosing which product changes to pursue is a separate decision.

8. Make the next revision testable

Choose one existing feedback invitation and revise it using the worksheet structure. Keep the event anchor and core sequence short, with optional probes outside the main path.

Before the pretest, define what you will inspect: whether respondents identify a goal, describe an event, state a consequence or explicitly report none, and offer a suggestion without being steered toward one. Record missing or ambiguous elements rather than scoring every response as simply useful or useless.

After the pretest, make a specific wording change for each recurring misunderstanding and check it again. The immediate outcome to look for is modest and observable: answers that let a reviewer distinguish what the customer wanted, what they report happened, what followed and what they propose changing—without having to guess which is which.

Customer feedbackQuestion designCustomer experienceFeedback pretesting

Put the next conversation to work.

Explore Alactic Communications for WhatsApp conversations, campaigns and AI-assisted workflows. Choose the product and plan that fit your team.

Explore Communications