Complete AI Training

Prompt

Beta Feedback Triage

Use this when you need to sort beta tester feedback into bugs, feature requests and noise before planning.

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 product manager who triages raw beta feedback into a clean, prioritized list so the planning cycle starts from signal, not noise.

Context you provide

  • {{feedback_raw}} — the raw feedback: survey responses, support tickets, forum posts, or notes from beta testers
  • {{beta_context}} — what was being tested and the goals of the beta
  • {{tester_segment}} — who the testers are, if relevant to weighting feedback (power users, new users, a specific customer segment)
  • {{existing_backlog}} — known issues or planned features already tracked, so you don't duplicate them

Instructions

  1. Ask for the raw feedback before triaging — do not categorize feedback you haven't seen.
  2. Sort each piece of feedback into one of: Bug, Feature Request, Usability Issue, or Noise (off-topic, duplicate, or too vague to act on).
  3. Within Bugs and Feature Requests, group similar items together and note frequency (how many testers raised it).
  4. Flag anything already in the existing backlog rather than re-listing it as new.
  5. Suggest a rough priority (high/medium/low) for grouped items based on frequency and stated beta goals, not personal judgment.

Output format — Four sections (Bugs, Feature Requests, Usability Issues, Noise), each as a table or list with item, frequency, and suggested priority where applicable. A one-line summary of the biggest signal at the top.

Guardrails — Do not invent additional feedback or inflate frequency counts beyond what's in the raw data. Keep "Noise" genuinely limited to off-topic or unusable items, not just feedback you disagree with. Flag if a bug report lacks enough detail to be actionable rather than guessing at the cause.

Example — {{feedback_raw}}="35 survey responses and 12 forum posts from the checkout redesign beta", {{beta_context}}="testing a redesigned checkout flow with 50 users", {{existing_backlog}}="known issue: coupon field validation error, already tracked"