A discovery call already contains everything needed for a proposal — the prospect said their pain points, budget signals, and success criteria out loud. Fathom captures that conversation as a searchable transcript and summary, Notion AI structures it into a reusable proposal outline, and ChatGPT drafts the actual language: executive summary, pain-point validation, proposed solution, and next steps. The job here isn't writing from scratch — it's extraction and structuring, which is exactly what AI does fastest, and it's the difference between a proposal going out the same day as the call versus three days later once someone finally gets around to writing it.
The three-step workflow
Step 1 — capture what was actually said. Fathom records and transcribes the call for free, highlighting objections, next steps, and follow-up commitments automatically, and syncs directly to HubSpot or Salesforce with one click — the raw material the rest of this workflow depends on.
Step 2 — structure it into a proposal shape. Notion AI takes the call transcript or summary and organizes it into a standard proposal skeleton — problem statement, proposed approach, pricing, timeline — so the next step is filling in a known structure rather than staring at a raw transcript deciding where to start.
Step 3 — draft the actual proposal language. ChatGPT turns that structured outline into client-ready prose: an executive summary that reflects the prospect's own stated priorities, a solution section that maps directly to what they said they needed, and a clear next-steps close.
Why this is a different job than proposal software
Our existing proposal and e-signature guide covers PandaDoc, Proposify, and HoneyBook — the *software* a finished proposal gets sent and signed through. This guide is scoped to the step before that: turning a raw sales conversation into the actual written content of the proposal in the first place, which most teams still do manually even when they already own proposal software.
Quick comparison
| Tool | Role in the workflow | Starting price | Job |
|---|---|---|---|
| Fathom | Call recording → transcript | Free | Capturing what was actually said |
| Notion AI | Transcript → proposal structure | $10/mo | Organizing it into a known shape |
| ChatGPT | Structure → finished draft | Freemium | Writing client-ready language |
How to actually run this same-day
Run Fathom during the call itself, so the transcript exists the moment the call ends. Feed the transcript into Notion AI immediately afterward while the context is fresh — this step takes minutes, not hours. Finish with ChatGPT for the actual prose, then have a human do a final pass for accuracy and tone before it goes out. Same-day proposals meaningfully outperform proposals sent days later, since buying momentum from the call itself is still active.
FAQ
Is it accurate to build a proposal directly from a call transcript, or does it need heavy editing? Treat the AI-generated draft as a strong first pass, not a final document — a human should verify pricing accuracy, confirm any commitments implied in the language, and adjust tone before sending, but the structural and drafting work genuinely doesn't need to start from a blank page.
Does this replace the need for proposal software like PandaDoc? No — this workflow produces the written content; you still need proposal software (or a simple document) to format, send, track opens, and collect a signature. The two are complementary steps, not competing tools.
How much of the call needs to be transcribed accurately for this to work? The parts covering pain points, budget signals, and success criteria matter most — a transcript with minor accuracy gaps elsewhere still supports a strong proposal draft as long as those core sections came through clearly.
Can this same workflow work for a proposal following multiple calls, not just one? Yes — feed Notion AI multiple call transcripts or summaries and ask it to synthesize across them into one structured outline before the ChatGPT drafting step, which matters more for longer, multi-stakeholder sales cycles.
Related guides
- Best AI Proposal and E-Signature Tool in 2026
- Best AI Tool for Auto-Logging Sales Calls Into Your CRM in 2026
- Best AI Tool for Writing Sales Follow-Up Emails After a Call in 2026
- Explore Business & Sales tools
*Ratings and pricing reviewed monthly. Last updated August 2026.*
Bogdex · Founder & editor, woska
Bogdex builds and curates woska, testing AI tools against real workflows to judge which ones actually save time rather than which have the longest feature list.