From coaching transcripts to an ICF-informed feedback report
How I turned an executive coach’s Granola transcripts into an ICF-informed feedback workflow, with session analysis and a short weekly report.
By Mollie Amkraut Mueller · · 2 min read
An executive coach I worked with wanted to use AI to evaluate her own coaching.
She was already saving session transcripts in Granola. We used those transcripts to build a Coaching Feedback Analyzer: a prompt that reviewed her sessions against an ICF-informed coaching framework.
The analysis focused on things she wanted to understand about her practice:
- Her talk-to-listen ratio.
- The percentage of her questions that were open-ended.
- Whether she was moving into advising rather than staying with coaching.
We ran it together on a real session in Granola. Her reaction helped shape the tool: she liked the quantitative feedback and was less interested in a long narrative summary.
That gave us a clear direction—make the review concise, with information she could quickly examine.
We started with a prompt she could run manually after any session she wanted to review. She tested it on her own sessions, reviewed the results and refined the prompt.
Once she was comfortable with the output, we automated the workflow into a short weekly coaching report.
The finished workflow turned transcripts she was already collecting into a regular source of feedback on her own practice. She could review a particular session when she wanted to, and receive a weekly report without having to initiate the analysis each time.
If you’re exploring how AI could fit into your own practice, I share practical workflows, tools and examples from projects like this in my newsletter, Watch Me AI.