As of 13 August 2026, AI can create a monthly summary of your customer feedback.

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 feedback, survey or complaints system and export the records for the month, including dates, ratings, comments, channel and any existing category fields.
   2. Remove unnecessary names, email addresses, phone numbers, order numbers and other personal data, then save the export with the reporting month and source recorded.
   3. Gather the definitions for each score, the NPS calculation if relevant, your existing feedback categories, and the previous month's export if you want a comparison.
   4. Paste the prompt into an approved chatbot, replace each bracketed slot, and attach or paste the redacted export and any previous-month data.
   5. Ask the chatbot to produce the summary and audit table, then compare every record count, score calculation and quoted comment with the source export.
   6. Check each proposed theme against the underlying comments and change or remove themes that combine different issues or imply causes not stated by customers.
   7. Send the checked summary to the relevant customer-service or product owner, keeping the original export and the final version together for the next month's comparison.

   Prompt

   Create a monthly customer-feedback summary from the data below.
   
   Reporting month: [MONTH AND YEAR]
   Business or team: [BUSINESS OR TEAM]
   Feedback source and definitions: [SOURCE, SCORE SCALE, NPS DEFINITION IF USED]
   Business priorities or categories: [PRIORITIES OR CATEGORIES]
   Feedback data:
   [PASTE A CSV EXPORT, TABLE, OR REDACTED CUSTOMER COMMENTS HERE]
   
   Use only the supplied data. Do not invent comments, customers, scores, causes, trends or actions. Treat blank, duplicated or unclear records explicitly and separate them from valid records. Keep personal data to the minimum needed and do not identify individual customers.
   
   Produce:
   1. A short executive summary in plain British English.
   2. The number of feedback records included, excluded and duplicated, with the reason for each exclusion.
   3. The relevant score results, including the calculation method and any missing values.
   4. The five strongest recurring themes, with record counts or proportions calculated from the supplied data and a short description of the evidence for each.
   5. Positive feedback themes and representative short quotes, clearly labelled as quotes.
   6. Complaint or negative-feedback themes, representative short quotes, and the customer impact described in the data.
   7. Any change from the previous month, but only if previous-month data is supplied; otherwise say that a comparison cannot be made.
   8. Questions, contradictions and possible data-quality problems that a person must resolve.
   9. A table of suggested follow-up actions, marked as suggestions rather than facts, with the evidence behind each suggestion.
   
   At the end, provide an audit table showing how every total and theme count can be checked against the supplied rows. Do not present correlation as causation and do not claim that the feedback represents all customers unless the data proves that.

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