win-loss
Skill3 starsChanged 31 days agoMITView on GitHub
Analyses sales call transcripts (and churn or renewal calls) to find why deals are won, lost, kept or churned. Reads each call across six dimensions (product, messaging, sales process, pricing, competition, customer context), pulls verbatim quotes with the speaker named, counts how often each pattern appears, and writes findings with recommendations. Works with any recorder export or pasted text. Triggers: "win/loss analysis", "why do we lose", "why do we win", "analyse these sales calls", "lost deals", "churn analysis", "objection patterns". Feeds /positioning and /messaging. Not for general meeting summaries or competitor research (use /competitor-research).
3 other projects use a version of this.
Adopted by
Most recently taken up by
- matteotitta/repo-structure-agencyword for word, first seen Nov 8, 2026
- jp-solumhealth/jpstacktheir own version, first seen Oct 14, 2026
- calven-ai/marketing-as-codetheir own version, first seen Oct 7, 2026
Other versions3 projects use a version of this.
Word for word · 1
- matteotitta/repo-structure-agency★ 1100% same
Edited · 0
None found.
Wrote their own version · 2
- calven-ai/marketing-as-code★ 10same name
- jp-solumhealth/jpstack★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository10 other files it gives its agents.
- brand-kitSkill
- company-contextSkill
- competitor-researchSkill
- expert-povSkill
- funnel-strategySkill
- goalsSkill
- icp-researchSkill
- messagingSkill
- positioningSkill
- tov-guidelinesSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 5 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 21
After 5 or more closed deals with recordings or notes.
- Distinctive: line 22
Before positioning or messaging work, to replace opinions about objections with evidence.
- Distinctive: line 23
When win rate drops, or a competitor starts showing up in losses.
- Very specific: line 26
Don't use it for one-off meeting notes, for researching competitors from the outside (/competitor-research), or for picking an ICP on its …
- Very specific: line 34
Paths: Write to context/win-loss/MMYY-win-loss.md.
- Distinctive: line 57
Read the strategy file: if one exists, and apply its locked definitions.
- Distinctive: line 75
Every finding has a frequency and at least one verbatim quote with a speaker role.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Win-loss
Turn sales calls into evidence about why people buy and why they don't. The quotes are the main product; the summary is secondary. Every pattern you report has to trace back to something a person said on a call.
When to use it
- After 5 or more closed deals with recordings or notes.
- Before positioning or messaging work, to replace opinions about objections with evidence.
- When win rate drops, or a competitor starts showing up in losses.
- Monthly, as a standing review for a team with steady deal flow.
Don't use it for one-off meeting notes, for researching competitors from the outside (/competitor-research), or for picking an ICP on its own (feed the findings into /icp-research).
Inputs
Required: transcripts, each with the customer (or a code name) and the outcome: win, loss, retention or churn. If the outcome isn't given, ask. Only infer it when the call itself states it.
Optional: deal size and stage, competitor names to watch for, context/icp/ and context/competitors/ for cross-referencing.
Paths. Write to context/win-loss/MMYY-win-loss.md. Keep a short context/win-loss/win-loss-strategy.md across runs: decisions made, definitions agreed, things to stop doing. Read it first on every run. Agency or consultant? Swap context/ for clients/<client>/context/ throughout.
Tools. None needed beyond the transcripts. If a call recorder or CRM is connected, pull from it; otherwise the user pastes or uploads files.
Before you analyse
- Normalise the format. Recorder exports, subtitle files, JSON and pasted text all become one shape: speaker-attributed turns, with timestamps where they exist.
- Redact personal data. Mask end-customer names, emails, phone and account numbers. Keep roles, company type and deal context.
- Bind every claim to a quote. A finding without a quote and a named speaker role doesn't ship.
The six dimensions
| Dimension | Win signals | Loss signals |
| Product | "Exactly what we need" | "It doesn't do X" |
| Messaging | "Now I see why this matters" | "What does it actually do?" |
| Sales process | "You understood our problem" | "The demo missed our use case" |
| Pricing | "Fair for what we get" | "Over budget", "too expensive" |
| Competition | "Chose you over Y" | "Going with Y" |
| Customer context | "We need this by Q2" | "No rush, maybe next year" |
Steps
- Read the strategy file if one exists, and apply its locked definitions.
- Process each call. Classify the outcome, tag speakers (buyer, champion, rep), and pull quotes against the six dimensions. Produces: a per-call evidence table.
- Aggregate by outcome. Group wins, losses, retention and churn. Count how many calls support each pattern ("4 of 7 losses").
- Rank patterns. Three or more mentions is a pattern. Two is emerging. One is a single mention, labelled as unconfirmed.
- Cross-reference by segment, deal size or competitor where the data allows. Look across at least three cuts, not just competitor.
- Separate product gaps from sales execution. They go to different teams.
- Write insights. For each: the pattern, the evidence with frequency and quotes, the confidence, and what to do about it.
- Update the strategy file with new decisions and definitions.
- Hand off to
/positioningand/messaging.
Output shape
Open with a 3-line TLDR and the sample size, stated once. Then: an executive summary, top win drivers, top loss drivers, objection themes with suggested responses, competitor mentions, a per-dimension breakdown, recommendations split into product, messaging and sales, and an appendix of quotes by call.
Confidence: high is 3 or more consistent calls, medium is 2 or strong indirect signals, low is a single mention.
What good looks like
- Every finding has a frequency and at least one verbatim quote with a speaker role.
- Patterns are described as "associated with", not "caused by".
- Small samples are flagged clearly. Under 5 wins and 5 losses, findings are directional.
- Dimensions with no data say "not discussed in transcripts".
- A seller could use the objection section on their next call.
Mistakes to avoid
- Summarising what buyers said instead of quoting them.
- Treating one memorable call as a pattern.
- Mixing up what the rep said with what the buyer said.
- Reading cause into correlation ("long cycles lose deals").
- Keeping customer names or emails in the saved file.
Chain
Step 4 of the suggested chain: /company-context → /icp-research → /competitor-research → win-loss → /expert-pov → /tov-guidelines → /brand-kit → /positioning → /messaging → /funnel-strategy → /goals.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

