# Win-Loss Analysis

How to turn deal outcomes into evidence about why you win and lose, and against whom.

## Sources and Their Bias

| Source | What it's good for | Bias to correct for |
|--------|-------------------|---------------------|
| CRM closed-lost reasons | Volume, trends over time | Rep-entered, often at close, skewed toward "price," "timing," and "no decision." Rarely names a product gap. |
| CRM closed-won notes | Which segments and competitors you beat | Thin. Reps rarely document why they won. |
| Call transcripts | The buyer's own words mid-deal | Shows what was said, not what decided it. Late-stage calls matter more than discovery. |
| Buyer interviews (won and lost) | The actual decision and who made it | Small samples. Lost buyers are polite, so ask for specifics. |
| Churned-customer notes | Gaps that showed up after the sale | Not a loss to a competitor unless they switched. Keep separate. |

When sources disagree, prefer the buyer's words over the rep's summary. "Price" in the CRM often turns out to be "didn't see enough value for the price" in the interview.

## Coding Scheme

Code each deal with one **primary reason**, optional secondary reasons, a supporting quote or note, and the source.

| Code | Use when |
|------|----------|
| Product fit | A missing or weaker capability decided it. Name the capability. |
| Price / value | Total cost or perceived value decided it. Note whether it was absolute budget or value-for-price. |
| Onboarding / time to value | Implementation effort, time, or services decided it. |
| Trust / risk | Security, references, company size, or vendor risk decided it. |
| Relationship / process | Incumbent relationship, champion left, procurement, or a better sales process. |
| Timing / no decision | The buyer didn't buy from anyone. Not a competitive loss. |

Rules:
- **One primary reason per deal**, so counts add up to the number of deals.
- **No reason without a source.** If a deal has only a one-word CRM reason, code it but mark it low-confidence.
- **Name the competitor** that won, or "no decision" or "built in-house." Don't infer it from who was mentioned once.

## Interview Questions

Ask within 2-4 weeks of the decision. Have someone other than the deal's rep do it if possible.

- Walk me through how the evaluation started. What triggered it?
- Who else did you look at, and how did the shortlist form?
- What were the 2-3 criteria that mattered most by the end?
- Where did we come out ahead? Where did we fall short?
- Was there a moment the decision tipped? What happened?
- If we'd changed one thing, would it have changed the outcome?

For lost deals, push past "price": "What would you have needed to see to choose us at that price?" Avoid "If we'd been the same price, would you have chosen us?" Polite buyers say yes.

## Reading the Results

- **Segment before counting**: by competitor, segment, deal size, and the stage the deal was lost at.
- **Report counts, not percentages, at small sizes.** "6 of 9" tells the reader how much to trust it. "67%" hides it.
- **Under ~5 deals in a slice, it's a signal to watch.** Don't put it on a battle card as a finding yet.
- **Separate competitive losses from no-decision.** Losing to inaction calls for different fixes than losing to a competitor.
- **Look at wins too.** The reasons you win against a competitor are what reps should lead with against them.

## Report Format

1. **Scope**: date range, number of deals (won / lost / no decision), sources used
2. **Headline findings**: 3-5, each with a count, a quote, and the segment it applies to
3. **By competitor**: win rate where known, top reasons won, top reasons lost
4. **Confidence notes**: which findings rest on thin or one-sided data
5. **Actions**: battle card updates, new or revised objection responses, discovery questions, product gaps

## Feeding Other Assets

| Finding | Goes to |
|---------|---------|
| Why we beat a competitor | Battle card "Where we win, and for whom" |
| Why we lose to a competitor | Battle card "When to walk away" or "Where they're strong" |
| Recurring objection | Objection library, with the response that worked |
| Criteria that decided deals | Discovery questions |
| Missing capability, repeated | Product team, with deal count and revenue lost |