Win-Loss Analysis Synthesizer

Every closed deal, won or lost, is free market research that most firms throw away. This skill synthesizes win-loss inputs into the recurring reasons deals go each way, separating what buyers say from what actually drove the decision, so the firm fixes the real causes instead of guessing.

1. The deal outcomes and any interview notes, debriefs, or buyer feedback.
2. The deals in scope and their basic facts (segment, competitor, size).
3. What the team believes is happening, so beliefs can be tested against evidence.

1. Separate stated reasons from root causes

Buyers rationalize. "Too expensive" often masks weak value framing or a missing champion. Record the stated reason and the likely root cause, and flag which is which.

2. Find the patterns, not the anecdotes

Cluster reasons across deals. A single dramatic loss teaches less than a quiet pattern across ten deals. Segment patterns by deal type, competitor, and stage where lost.

3. Diagnose won deals too

Why you win is as actionable as why you lose. Wins reveal the messaging, proof, and buyer type where you are strong, which should shape targeting.

4. Distinguish fixable from structural

Some losses are fit problems (wrong ICP) and some are execution problems (slow follow-up, weak discovery). The response differs, so label each.

5. Turn patterns into changes

Each significant pattern gets a specific recommended change to messaging, process, targeting, or product.

ALWAYS use:

# Win-Loss Synthesis | [Period / Deal Set]
## Why we win (patterns and where we are strong)
## Why we lose (stated reason vs root cause, clustered)
## Patterns by segment / competitor / stage
## Fixable vs structural
## Recommended changes (ranked by impact)

• Taking buyer-stated reasons at face value.
• Over-indexing on one memorable loss.
• Analyzing only losses and ignoring what wins teach.
• Producing observations with no recommended change.

Across ten losses, the pattern is deals stalling after a strong demo with no economic buyer engaged. Root cause is single-threading in discovery, and the recommended change is a required multi-threading step before demo.