Win/Loss Analyzer

You know what you lost. You don't know why.

Win/Loss Analyzer is an OpenClaw skill that runs an end-to-end win/loss program: guided interview scripts for win, loss, no-decision, and churn calls; transcript coding with a six-category reason taxonomy; cross-deal aggregation with frequency × dollar-value analysis; and a quarterly report with cited quotes and recommended actions for sales, product, and marketing. The whole pipeline replaces a $5–15K/quarter consulting engagement.

What it does

• Guides an interviewer through a structured 20–30 minute script with branching probes (win, loss, no-decision, or churn types)
• Codes each prospect turn with reason categories (price, fit, competitor, sales experience, timing, trust) and sentiment
• Aggregates across 5–25 coded deals to produce frequency heatmaps and competitor mention counts
• Renders a distributable quarterly report with top themes, supporting quotes, and specific recommended actions per team

What's in the package

All files listed below are bundled as fenced code blocks inside SKILL.md — there's no separate folder to install. Copy each code block into its own file using the path shown above it.

• SKILL.md — the full workflow your agent follows.
• scripts/run_interview.py — interactive CLI for guided win/loss interview sessions.
• scripts/code_transcript.py — reason taxonomy coder for raw transcripts or interview logs.
• scripts/aggregate_themes.py — cross-deal aggregation with dollar-weighting support.
• scripts/render_quarterly.py — quarterly report renderer in distributable markdown.
• data/interview-scripts.yaml — pre-built question sets for all four interview types, plus the full reason taxonomy.
• scripts/examples/ — sample transcripts, coded JSONs, aggregated output, and a quarterly report.

How it ships

Drop SKILL.md into ~/.openclaw/skills/win-loss-analyzer/. Restart OpenClaw. Run /win-loss-analyzer or say 'start a loss interview for [Deal Name].'

Questions: hackedcubeapps@gmail.com (response within 48 hours).