/compete

Win/Loss Analysis Reference

Purpose: Systematic post-decision interview program that surfaces the real reasons deals were won or lost — not the reasons the rep wrote in the CRM. Win/loss is the highest-signal competitive intelligence input because it is grounded in actual buyer decisions, not analyst opinion or competitor marketing. A formal program yields 15-30% win-rate improvement.

Scope Boundary

• compete winloss: post-decision buyer interviews, theme extraction, cadence design, integration with CRM and battle cards.
• compete matrix (default, elsewhere): feature comparison — descriptive, not behavioral. Win/loss explains why features matter, not what they are.
• compete swot (elsewhere): strategic SWOT — win/loss data feeds SWOT but is not SWOT itself.
• compete positioning (elsewhere): positioning maps — win/loss surfaces positioning gaps; positioning resolves them.
• compete battle (elsewhere): battle cards consume win/loss themes as their evidence layer. Do not author "why we win/lose" sections from opinion.
• voice (elsewhere): existing-customer feedback and NPS. Voice is post-purchase satisfaction; win/loss is purchase-decision rationale. Different population, different question set.
• field (elsewhere): discovery-phase user research and persona work. Field designs the methodology; Compete owns the win/loss program structure and competitive lens.
• echo demand: synthetic-user assumption challenge. Echo[demand] is hypothesis stress-testing without real interviews — never substitute for real win/loss.

Workflow

DESIGN    →  define segments (deal-size x vertical x role x outcome)
          →  draft interview script; choose third-party interviewer

RECRUIT   →  invite within 2-6 weeks of decision (sweet spot)
          →  target 30-40% response rate; offer incentive ($100-250)

INTERVIEW →  30-45 min recorded calls; let buyer narrate timeline
          →  cover: trigger, evaluation set, decision criteria, deciding factor

CODE      →  tag each interview by themes; require 3+ mentions to elevate
          →  separate signal from one-off complaints

SYNTHESIZE → quarterly report: themes by segment, trend deltas, action items
           → feed battle cards (objections, why-win/lose), Spark (gaps)

CADENCE   →  quarterly cycle minimum; monthly for high-velocity SMB sales
          →  CRM-integrate findings; close the loop with sales + product

Segmentation Axes

Minimum viable program: segment by outcome x deal-size x competitor. Add vertical and role once volume supports cross-tabulation (typically 40+ interviews/quarter).

Interview Script Skeleton

Question rule: never lead. "Was pricing the issue?" yields useless yes/no; "What were the top 3 factors?" yields ranked truth.

Theme Coding Workflow

1. Transcribe every interview (Otter, Fireflies, Gong).
2. First-pass code: tag with predefined themes (price, features, integration, trust, timing, champion-loss, ROI-clarity).
3. Second-pass code: tag emergent themes not in the predefined set.
4. Elevation rule: a theme requires 3+ mentions across interviews before it qualifies as a finding. One-off complaints are noise.
5. Cross-segment analysis: a theme that appears in one segment but not others is a segment-specific finding, not a global one.
6. Quote selection: for each elevated theme, pick 2-3 verbatim quotes as evidence (anonymized).

Signal-to-Noise Filtering

Noise filters: complaints from buyers who never reached late-stage evaluation, blame-shifting language ("my CFO killed it"), and post-rationalization more than 8 weeks after decision.

Third-Party Interviewer Benefits

Recommendation: use third-party for loss interviews and high-stakes wins; internal acceptable for routine wins where confirmation is the goal.

Cadence Design

Quarterly is the default minimum for B2B SaaS. Below 20 competitive deals/quarter, run ad hoc but never skip — even 5 interviews/quarter beats zero.

CRM Integration

Integration rule: findings that do not flow back into CRM and battle cards are research theater. Close the loop or stop running the program.

Anti-Patterns

• Rep-authored loss reasons — reps write "price" or "features" because it protects relationships and pipeline; buyer reality differs in 60-70% of cases.
• No segmentation — pooled win/loss findings hide that SMB lost on price while enterprise lost on integration. Segment or learn nothing.
• Interviewing only losses — wins reveal what to protect; losses alone bias the program toward defensive thinking.
• Interviewing too late — beyond 8 weeks, buyer recall degrades and post-rationalization dominates. Interview within 2-6 weeks.
• One-and-done program — quarterly cadence is the minimum; one-off projects cannot detect trends or measure improvement.
• No theme-elevation rule — promoting a single quote to a finding produces strategic whiplash. Require 3+ mentions.
• Findings without action — if win/loss does not change battle cards, roadmap, or pricing, the program will be defunded.
• Interviewer without training — untrained interviewers ask leading questions and miss the deciding-factor probe.
• Treating "price" at face value — price is the most-cited and least-real loss reason. Probe for ROI clarity, champion strength, and timing instead.

Handoff

• To Voice: pull post-purchase satisfaction quotes from existing customers as supporting evidence for "why we win" themes.
• To Field: hand off interview script design, persona validation, and qualitative coding methodology refinement.
• To Magi: when win/loss reveals a structural shift (segment moving, category re-defining), escalate for strategic simulation.
• To Spark: feature gaps that appear in 3+ loss interviews route as feature ideas with competitive and buyer evidence attached.
• To Growth: positioning, brand-authority, or first-touch trust gaps route to Growth for messaging and SEO/GEO response.
• To battle (compete): every quarter's themes feed the "why we win / why we lose / objections" sections of battle cards.
• To Lore: recurring win/loss patterns that hold across 3+ quarters become institutional knowledge.