AI Customer Feedback Analysis Workflow

5 steps to a reviewable draft

1. 01

   Set the question

   Decide what the feedback review should inform, such as a service change, onboarding fix or pricing conversation.

   Result: A specific decision question.

2. 02

   Prepare the source set

   Keep each comment, date, customer context and source channel attached before grouping themes.

   Result: A feedback set with traceable excerpts.

3. 03

   Group the feedback

   Ask for themes, direct excerpts and contrary evidence instead of an invented customer consensus.

   Result: A feedback sheet with evidence beside each theme.

   Your prompt

   Copy prompt

   Analyze this customer feedback for the question: {{decision_question}}. Source excerpts with date and context: {{feedback}}. Return themes, representative excerpts, contrary or outlier feedback, frequency notes only when the data supports them, and questions to investigate. Do not invent sentiment scores, customer segments, demand, churn causes or product priorities.

   Example input

   Question: why do customers delay onboarding? Feedback includes two comments about missing setup instructions and one comment about internal approval timing.

   Example output

   Theme: setup instructions. Counterexample: one customer cites internal approval timing. Follow-up: review the onboarding handoff.

4. 04

   Check representation

   Verify that the strongest theme does not hide an important contrary comment, small sample or change in customer context.

   Result: A balanced synthesis that the owner can inspect.

5. 05

   Decide what to learn next

   Pick one follow-up, owner and deadline rather than assigning a product change directly from the summary.

   Result: A feedback-informed next action.