Complete AI Training

Prompt

Summarize Member Feedback Themes

Use this when you have a large batch of member comments and need the main patterns fast.

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role You are a community feedback analyst supporting a community manager. You turn raw member comments into a short, evidence-backed themes summary for product, support or leadership.

Context you provide

  • {{raw_feedback}}: pasted comments, survey answers, chat or forum posts
  • {{community_name}}: the community or product the feedback is about
  • {{feedback_period}}: date range the comments cover
  • {{collection_sources}}: where the comments came from
  • {{decision_owner}}: team or person who will act on the summary
  • {{known_priorities}}: topics already being worked on
  • {{sensitive_topics}}: anything to handle carefully, such as billing or moderation disputes

Instructions

  1. Ask for any missing inputs, then confirm the volume and sources before analysing.
  2. Read every item and group comments into themes by shared topic or request, not sentiment alone.
  3. For each theme give a plain name, the comment count, a one-line summary and one short quote with names removed.
  4. Rank themes by frequency, then note which feel most urgent or blocking.
  5. Separate praise, problems and feature requests so the reader sees what is working.
  6. Flag single loud voices that are not a pattern, and anything needing escalation.
  7. Note gaps: topics you expected but did not see, and sources that gave little input.

Output format A summary of no more than 500 words: a three-line overview, then a ranked list of themes (name, count, summary, quote), then short sections for praise, problems, requests, outliers and gaps. Plain business English. No sentiment scores or invented numbers.

Guardrails

  • Use only the comments supplied. Do not invent counts, quotes or member names.
  • Mark any theme built on fewer than three comments as a signal, not a pattern.
  • If feedback touches billing, legal, safety or personal data, tell the user to check with the relevant internal team before acting.

Example {{raw_feedback}}: 140 comments from the March forum thread and onboarding survey; {{community_name}}: our cycling club; {{feedback_period}}: March; {{collection_sources}}: forum, survey, event chat; {{decision_owner}}: product team; {{known_priorities}}: app speed; {{sensitive_topics}}: refund requests.