As of 13 August 2026, AI can create a customer feedback report.

What to actually do

1. Hand it to a person

   The route this page recommends

   A person who owns the outcome does this end to end, worth it when the failure is dear.

2. Use a tool built for this

   Second choice

3. Do it yourself

   The distant third

   A chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.

   How to actually do it

   1. Open the survey, review, complaint or support export and remove names, email addresses, telephone numbers, order references and any other unnecessary personal data.
   2. Gather the reporting period, response channel, question wording, rating scale, relevant customer segments, product names and the business decision the report will support.
   3. Paste the business context and anonymised feedback into the prompt, keeping the original rating values and response text intact.
   4. Ask the chatbot to produce the report and show all NPS or other numerical calculations rather than presenting unexplained totals.
   5. Compare every reported count, percentage, rating and quotation against the source export, and correct any missing, duplicated or misclassified responses.
   6. Check that each theme is supported by enough relevant responses for your context, and mark conclusions based on sparse or biased feedback as hypotheses.
   7. Remove any unnecessary personal information from the draft, send safety, discrimination, privacy, fraud or legal concerns to a responsible person, and share the finished report with named owners for the agreed actions.

   Prompt

   Create a customer feedback report for my UK business using only the data and context I provide below.
   
   Business context:
   [describe the business, customers, products or services, reporting period and the decision this report will support]
   
   Feedback data:
   [paste anonymised survey responses, review text, complaints, support notes or an export]
   
   If the data includes ratings from 0 to 10, calculate the response count and NPS where the necessary ratings are present. Show the calculation, state the denominator and identify missing or unusable ratings. Do not invent responses, ratings, customer details, dates, causes or trends.
   
   Produce the report with these sections:
   1. Executive summary
   2. Data scope and limitations
   3. Overall satisfaction and NPS findings, if the data supports them
   4. Main positive themes, with the number of source responses where you can verify it
   5. Main problems or negative themes, with supporting quotations or response references
   6. Differences between customer groups, products or channels only where the data supports a comparison
   7. Urgent issues that need a human to investigate, including any safety, discrimination, privacy, fraud or legal concern
   8. Practical actions, separating evidence-based actions from hypotheses to test
   9. A short table of each finding, its evidence, confidence and suggested owner
   
   Keep customer quotations anonymous. Preserve the original meaning and flag unclear, contradictory or duplicated responses. Label every interpretation as an interpretation rather than a fact. End with a list of checks I must complete before sharing the report. Use clear British English and a neutral tone.

   Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.