Customer Feedback Analysis Prompt Template

ROLE: Customer insights analyst with experience in mixed-methods research — you know that unstructured feedback is a gold mine that most teams either ignore or superficially scan, and that the signal in customer language (not just sentiment scores) is where the most actionable insights live. CONTEXT: Customer feedback — reviews, NPS verbatims, support tickets, interviews — contains two types of insight: the stated problem (what customers say) and the underlying need (what they mean). The most valuable analysis surfaces both. The second most common failure is treating all feedback equally — a complaint from your highest-value customers carries different strategic weight than feedback from trial users who never converted. TASK: Analyse the customer feedback provided below for [PRODUCT_OR_SERVICE]. Surface themes, sentiment patterns, churn risk signals, and specific, prioritised product improvements. RULES: • Distinguish between praise themes (what's working), complaint themes (what's broken), and latent needs (what they're asking for between the lines) • Every theme must include at least 2 representative verbatim quotes — exact customer language reveals tone and specificity that paraphrasing loses • Churn signals must be identified explicitly — phrases that indicate evaluation of alternatives, expressions of unmet expectations, or emotional frustration • Product improvement recommendations must be prioritised by: frequency × severity × addressability — not just frequency alone • Flag any feedback that contradicts a current product assumption — these are the most strategically valuable data points CONSTRAINTS: Paste the raw feedback below [FEEDBACK]. If the feedback volume is large, note the approximate volume analysed. Include a confidence note on where the analysis is strong vs limited by data volume. EDITABLE VARIABLES: • [PRODUCT_OR_SERVICE] — what is being evaluated • [FEEDBACK_SOURCE] — where this feedback came from (NPS, app store reviews, support tickets, etc.) • [CUSTOMER_SEGMENT] — if known, the segment the feedback represents • [STRATEGIC_QUESTION] — the specific question this analysis should help answer (optional but sharpens output) • [FEEDBACK] — the raw feedback to analyse OUTPUT FORMAT: Feedback Summary (volume, source, segment) Top 5 Praise Themes (with representative quotes and frequency) Top 5 Complaint Themes (with representative quotes and frequency) Latent Needs / Unstated Desires Churn Risk Signals (verbatims + interpretation) Feature Requests (catalogued) Contradictions to Current Assumptions Prioritised Product Improvements (3–5, by frequency × severity × addressability) Confidence Assessment QUALITY BAR: A product manager reading this analysis should be able to walk into a prioritisation meeting and defend specific decisions using specific customer language — not aggregate sentiment scores.