Win/Loss Interview Analysis Prompt
Prompt
You are a revenue operations manager analyzing win/loss interview data. Interview data: [PASTE: Deal | Won/Lost | Competitor (if lost) | Customer's primary decision factor | Our strengths cited | Our weaknesses cited | Competitor strengths cited | What would have changed the outcome] Analyze: 1. Win themes — what do we win on consistently? (price / functionality / ease of use / support / relationships) 2. Loss themes — what do we lose on consistently? (feature gaps / price / implementation complexity / competitor relationship) 3. Competitive patterns — against which competitors do we win/lose most? What is the deciding factor in each matchup? 4. Decision factor ranking — what matters most to buyers in our category? Are we strong or weak on what matters? 5. Avoidable losses — how many losses were due to factors we could have changed? (discovery failure / late engagement / wrong stakeholder) Output: Win/loss analysis. Win and loss themes. Competitive matchup summary. Top 3 actionable findings for sales, product, and marketing.
Why it works
Synthesising multiple win/loss interviews reveals patterns that individual interviews can't — a theme that appears in 60% of lost deals is a strategic problem, while one that appears in 10% may be a sales execution outlier. Separating customer-cited reasons from interviewer interpretation ensures the analysis preserves the raw customer voice alongside the strategic conclusions. The product versus sales execution split is the most commercially important distinction, as these require completely different investment responses.
Watch out for
Win/loss interview analysis is subject to selection bias — customers who agree to interviews are typically either very happy (wins) or very principled about honest feedback (losses), and the majority of customers who decline to participate may have different perspectives. Flag the participation rate in the analysis and treat conclusions as directional rather than definitive until sample sizes are large enough to produce statistically reliable patterns.
Used by