Your Best Deals Hold the Answers – If You Know How to Read Them
Every closed deal – won or lost – contains signals about your positioning, your sales motion, and where competitors are gaining ground.
The problem is that extracting those signals manually takes weeks. CRM exports, transcript reviews, survey analysis, competitive research – by the time you finish, the insights are stale.
Claude handles the entire win/loss workflow end-to-end, pulling data from your CRM, analyzing call recordings, synthesizing customer feedback, and delivering a structured playbook your team can act on immediately.
How Claude Helps You
Claude runs a complete win/loss analysis by chaining together five specialized skills – each one handling a different intelligence source. It pulls your deal data, reads through call transcripts, aggregates customer feedback, mines competitor reviews, and identifies loss patterns. Then it synthesizes everything into a single actionable report.
To do this, Claude uses these skills:
- pipeline-review – Pulls deal data from any CRM and produces pipeline diagnostics with win/loss breakdowns (Gooseworks)
- meeting-insights-analyzer – Analyzes call transcripts for behavioral patterns, speaking ratios, and objection moments (ComposioHQ)
- voice-of-customer-synthesizer – Aggregates customer feedback from support tickets, NPS, surveys, and reviews into themed clusters (Gooseworks)
- review-intelligence-digest – Extracts recurring themes, objections, and proof points from G2, Capterra, and Trustpilot reviews (Gooseworks)
- battlecard-generator – Researches competitors across web, reviews, ads, and social to surface win/loss themes and positioning traps (Gooseworks)
What you provide: CRM access (or a CSV export), call transcripts from the past quarter, and the names of 1–3 competitors you lose deals to most often.
The Win/Loss Workflow
Step 1: Pull Deal Data from Your CRM
Claude connects to your CRM – whether that's Salesforce, HubSpot, Pipedrive, or a spreadsheet – using pipeline-review. It pulls every deal from the past 90 days and maps them to your pipeline stages.
From there, it calculates win rates by stage, source, deal size, and rep. If 60% of your losses happen between demo and proposal, that shows up immediately.
The output includes both an executive summary and a detailed diagnostic with stage-by-stage conversion data.
Step 2: Analyze Call Transcripts for Behavioral Patterns
Next, Claude reads through your sales call transcripts using meeting-insights-analyzer. It identifies patterns that correlate with wins and losses.
- Speaking ratios – are reps talking 70% of the time on lost deals vs. 40% on wins?
- Objection moments – where do prospects push back, and how do reps handle it?
- Conflict avoidance – are reps sidestepping pricing conversations or competitive questions?
Claude flags specific timestamps and quotes, so your sales manager can coach with real examples rather than gut feelings.
Step 3: Synthesize Customer Feedback
Claude aggregates feedback from every channel you have – support tickets, NPS responses, churn surveys, Slack messages, and email threads – using voice-of-customer-synthesizer.
It clusters feedback into themes, tags sentiment (positive, neutral, negative, critical), and identifies which themes appear most often in churned accounts vs. retained ones.
If 8 out of 12 lost deals mention "onboarding complexity" in their exit surveys, that pattern surfaces with exact quotes attached.
Step 4: Cross-Reference with Review Intelligence
Claude scrapes your G2, Capterra, and Trustpilot reviews – plus your competitors' reviews – using review-intelligence-digest.
It extracts five lenses of insight:
- Proof points – specific outcomes and metrics from 5-star reviews
- Pain language – the exact words prospects use to describe their problems
- Competitor complaints – what your competitors' customers hate most
- Feature gaps – what customers wish existed
- Switching signals – why customers leave one product for another
This tells you not just why you're losing, but what language to use when you start winning those deals back.
Step 5: Identify Competitive Loss Patterns
For every competitor you lose deals to, Claude runs battlecard-generator to build a detailed competitive profile.
It analyzes their website messaging, review sentiment, ad copy, and social presence. Then it maps your loss patterns against their strengths to answer the critical question: are you losing on product, positioning, or price?
Each battlecard includes landmine questions for reps, objection handlers, and positioning traps – the exact ammunition your team needs in competitive deals.
Step 6: Build the Win/Loss Report
Finally, Claude synthesizes everything from the previous five steps into a structured win/loss playbook. No skill needed here – Claude reads the outputs from each step and produces the final deliverable.
The report lands in your preferred format – a Notion page, Google Doc, or markdown file – organized by loss reason, competitor, and recommended action.
What You Walk Away With
- Win/loss breakdown by segment – conversion rates sliced by deal size, source, rep, and stage
- Call pattern analysis – speaking ratios, objection handling scores, and coaching moments with timestamps
- VoC theme report – clustered customer feedback with sentiment tags and churn correlation
- Competitive intelligence digest – review-mined insights for your product and top 3 competitors
- Battlecards per competitor – positioning traps, landmine questions, and objection handlers
- Actionable playbook – prioritized recommendations tied to specific loss patterns
Why This Matters
Win/loss analysis is the fastest way to improve close rates – but only if the insights reach your team while deals are still in motion. Running this workflow with Claude Code, Codex, or Goose turns a quarterly research project into something you can run every month, with fresher data and tighter feedback loops. The teams that operationalize win/loss analysis don't just understand why they lose – they stop losing for the same reasons twice.
Start Running Win/Loss Analysis with AI
Install a skill and run your first analysis today. Visit gooseworks.ai to get started.