All use cases
Sales Follow-up Drafting
AI-drafted follow-up emails after sales calls. 45 minutes down to 8 minutes per prospect.
The problem
Sales reps spending 45 minutes per prospect on follow-up emails after meetings. Quality wildly inconsistent across the team. Some reps writing novels, others sending two-line nothing-emails. Both losing deals.
The result
Follow-up time dropped from 45 minutes to 8 minutes per prospect. Email reply rate increased 22%. Reps started editing less over time as prompts improved.
The workflow
How the system actually runs.
- 01 Deal moves to "Meeting Completed" stage in HubSpot
- 02 Make pulls meeting notes and deal context from HubSpot
- 03 Claude drafts personalized follow-up email using context
- 04 Draft lands in rep's email as a pre-filled template
- 05 Rep reviews, edits if needed, and sends
Tools used
The stack behind it.
Claude APIMakeHubSpot
What worked
Why this setup held up.
- Including meeting notes in context made drafts actually relevant — not just generic templates
- Building persona-specific templates (CFO vs CTO vs Head of Ops) dramatically improved quality
- Reps could override the tone with a one-word flag: "formal" or "casual"
What did not
The friction to watch.
- Generic one-size-fits-all prompts produced generic emails — had to specialize per buyer persona
- Without meeting notes, the AI just rewrote the company boilerplate — useless
- Some reps resisted for two weeks until they saw reply rate data
Verdict
The short version.
Massive time saver. The key is feeding enough context — garbage in, garbage out. Meeting notes are non-negotiable input.