A Claude skill that reads your sales call transcripts from Google Drive, scores every deal on MEDDPICC, surfaces your top objections, flags at-risk pipeline, and delivers a polished weekly intelligence report — automatically.
Every week, this skill connects to a Google Drive folder containing your call transcripts, reads every one, and produces a structured Google Doc report that tells you exactly what happened across your pipeline and what to do about it.
It's designed for Account Executives who want more than a CRM snapshot — it gives you the conversation-level intelligence that lives inside calls but rarely makes it into Salesforce.
For every transcript in your folder, the skill extracts:
- MEDDPICC score (0–16) — each element scored individually so you know exactly where the gaps are, not just whether a deal is "qualified"
- Objections — verbatim, with notes on whether they were resolved, deflected, or left open
- Buying signals — classified as Strong / Moderate / Weak / Negative, with the specific moments that drove the rating
- Competitor mentions — who came up, how, and in what context
- Pricing & budget signals — what was said, by whom, and how it landed
- Action items & next steps — what you committed to, what the prospect committed to, and a flag if no next meeting was confirmed
- Product feedback — feature gaps, integration concerns, and things prospects wish the product did differently
- Coaching observations — one thing you did well, one specific thing to try differently next time
Across all your calls, the report surfaces:
- Top 5 objections ranked by frequency, with suggested talk tracks for each
- Competitor leaderboard — who's coming up most and how you're positioned against them
- Deals to prioritize — ranked by buying signal strength + MEDDPICC score combined
- At-risk flags — deals with low MEDDPICC scores, no confirmed next steps, or negative sentiment
- Product feedback themes — grouped across all calls, traceable back to source conversations
- Personal coaching summary — your pattern strengths and one focused area to work on next week
📊 Executive Summary
• 6 calls analyzed. Acme Corp and Vertex Health show strongest buying intent this week.
• "We already have a solution for that" came up in 4/6 calls — unresolved in 3 of them. Needs a talk track.
• No next steps confirmed on 2 calls (DataSync, Horizon Retail) — follow up today.
• Salesforce mentioned in 3 calls, all in "evaluating alongside" context.
🔥 Deals to Prioritize
Company | MEDDPICC | Signal | Next Step | Risk
Acme Corp | 13/16 | Strong | Legal review by Fri | —
Vertex Health | 11/16 | Strong | Champion intro to VP | Paper process unknown
⚠️ At-Risk Deals
• DataSync (4/16) — No economic buyer surfaced, no next steps confirmed
• Horizon Retail (6/16) — Went cold after pricing discussion
🚫 Top 5 Objections This Week
1. "We already have something for that" (4 calls, resolved 1/4)
→ Talk track: "That's common — most teams we work with had a point solution before us.
What's been the biggest gap you've noticed with what you have today?"
...
| Requirement | Details |
|---|---|
| Claude app | Cowork desktop app (where skills run) |
| Google Drive MCP | Must be connected in Cowork — the skill reads your transcript Docs and writes the report back to Drive |
| Transcript format | Google Docs in a shared Drive folder. Works with Gong exports, Otter.ai, Zoom auto-transcripts, and manual notes |
| Sales framework | Built around MEDDPICC — works best if your team already uses this, but useful even if you don't |
Download call-transcript-analyzer.skill from this repo (Releases tab, or directly from the root).
Open the Cowork desktop app, then either:
- Drag and drop the
.skillfile into the app, or - Click the skill file — it will prompt you to install
In Cowork, connect the Google Drive MCP if you haven't already:
- Go to Settings → Connectors
- Find Google Drive and click Connect
- Authorize with the Google account that has access to your transcript folder
Just talk to Claude naturally:
"Analyze my call transcripts in the 'Q2 Calls' folder and give me this week's report."
Claude will find the folder, read every transcript, and write the report as a new Google Doc in the same folder (or a "Weekly Reports" subfolder if one exists).
Once you've confirmed the skill works with your transcripts, you can schedule it to run automatically every Monday morning:
"Set up the call transcript analyzer to run every Monday at 8am on my Q2 Calls folder."
Claude will configure a recurring task. You'll get a fresh report waiting for you at the start of every week.
call-transcript-analyzer/
├── SKILL.md # Core skill instructions
└── references/
└── meddpicc.md # MEDDPICC scoring criteria (0–2 per element)
The skill is a plain Markdown file (SKILL.md) — no code, no build step. You can open it in any text editor and adjust:
- Sales framework — swap MEDDPICC for BANT or a custom framework by editing the scoring section
- Report sections — add, remove, or reorder sections in the "Write the report" step
- Coaching criteria — adjust what counts as good/bad talk ratio, question quality, etc.
- Folder behavior — change where the output report is saved
After editing, repackage by zipping the folder with a .skill extension and reinstalling.
This is a Claude skill — a Markdown file that gives Claude a specific set of instructions and context when triggered. When you ask Claude to analyze your calls, Cowork loads SKILL.md into Claude's context alongside the Google Drive MCP tools. Claude then:
- Lists all Google Docs in your specified folder
- Reads each transcript in full
- Runs the per-call analysis pass across all of them
- Aggregates findings into the weekly view
- Writes the final report as a new Google Doc via the Drive MCP
No data is stored outside of your Google Drive. No external servers are involved beyond Anthropic's Claude API and Google's Drive API.
PRs welcome. Useful directions:
- Support for additional transcript formats (Chorus, Fireflies, etc.)
- CRM-aware deal stage mapping (e.g., pull current stage from Salesforce to compare against transcript signals)
- Multi-AE team mode (aggregate across a team folder, not just one AE's calls)
- Slack digest output alongside the Google Doc
MIT — use freely, modify for your team, share with your org.