analyze-win-losslisted

# Analyze Win Loss Explain what the evidence supports about buying decisions and what the team should test next. Do not reduce a complex decision to one convenient reason code. ## Inputs And Scope Read `.agents/gtm-context.md` when available. Confirm: - decision period, segment, product, geography, and opportunity population; - outcome definitions for won, lost, no decision, disqualified, and unknown; - source types, response rates, missing records, and who supplied each explanation; - whether the task is a single-deal review, a cohort comparison, or a portfolio analysis. Preserve a stable opportunity identifier while removing unnecessary personal data. Do not merge similarly named accounts or double-count reopened opportunities. Read [the analysis methods](references/methods.md) for coding evidence, comparing cohorts, interviewing buyers, and handling small samples. ## Workflow 1. Audit outcome labels, duplicates, missing fields, reopened deals, and inconsistent dates before analysis. 2. Build a source map for each decision. Keep buyer evidence, seller interpretation, CRM fields, and product or commercial records distinct. 3. Reconstruct the buyer's decision: desired change, alternatives, criteria, process, stakeholders, decisive events, and final outcome. 4. Code contributing factors consistently across wins, losses, and no-decisions. Allow multiple factors per outcome. 5. Label each finding as `Verified`, `Inferred`, `Unknown`, or `Contradicted`, with a source and