Meeting Notes, Synthesis & Client Follow-ups

A consultant in a typical week sits in 8–15 client meetings. Without AI, you either write notes during the conversation (and miss subtext) or after (and forget half the detail). With modern AI transcription and synthesis, you fully participate in the meeting and still get better notes than you ever did manually.

This lesson is the most immediately actionable in the course — most consultants who adopt this workflow save 4–6 hours per week.

What You'll Learn

• The transcription tools that work for client meetings
• A four-step AI workflow: capture, summarize, structure, follow up
• How to draft client follow-up emails in 30 seconds
• How to build an institutional memory across recurring engagements

Choosing a Transcription Tool

The market in 2026 has matured. The main credible options:

• Otter.ai — strong general-purpose, easy to share, integrates with Zoom and Teams.
• Fireflies.ai — built for sales and consulting, automatic CRM logging, good speaker identification.
• Granola — local-first, very popular with consultants because it handles in-person meetings, Mac-only.
• Microsoft Teams "Intelligent Recap" — built into Teams, no third-party data sharing, ideal for enterprise consulting.
• Zoom AI Companion — equivalent for Zoom-first clients.

For client work, prefer the tool that comes with the meeting platform (Teams Recap or Zoom AI Companion) — there is no third-party data flow. For internal meetings or your own discovery work, Granola or Otter is fine.

Always disclose recording. Most jurisdictions require consent. A simple "I'm going to enable our AI notetaker — okay with you?" at the top of the call is sufficient and shows respect.

The Four-Step Workflow

Step 1: Capture

Start the AI notetaker before the call begins. Confirm verbally. Take only a handful of light handwritten notes for things the transcript will miss — body language, the partner you saw frown when a topic came up, the topic the CFO carefully avoided. These observations are your edge.

Step 2: Summarize

Most transcription tools generate an automatic summary. These are usually shallow. Replace them with a structured AI prompt instead.

Below is the transcript of a 60-minute client meeting between [client roles] and our consulting team. Produce: (1) a 5-bullet executive summary at C-level abstraction, (2) the 3 most important things [primary client contact] said, in their own words, (3) any decisions made (with who decided), (4) any commitments made by us, (5) any commitments made by the client, (6) open questions we owe an answer to, (7) topics that came up but were deferred. Output as markdown with clear headers.

Transcript: [paste]

The "commitments made by us" and "commitments made by the client" sections are gold. They are the single most common source of trust loss in consulting — you forget what you promised, the client remembers.

Step 3: Structure into Your Knowledge System

Drop the structured summary into wherever you keep engagement notes — Notion, OneNote, Obsidian, your firm's Confluence. Tag with the engagement, the date, and the participants.

For multi-month engagements, do a periodic synthesis prompt:

Below are 12 weekly meeting summaries from a 3-month engagement with ClientCo. Produce: (1) the evolving narrative — how the client's framing of the problem has changed over the engagement, (2) commitments we made and which are still open, (3) topics raised by the client that have not yet been answered, (4) any contradictions or tensions that have surfaced over time. This is for the partner's pre-read for the steerco.

This is impossible to do manually with the same rigor.

Step 4: Draft the Follow-up Email

Within 30 minutes of every client meeting, send a follow-up. AI makes this trivial.

Draft the follow-up email to [client contact, role] after the meeting summarized below. Tone: warm but professional, direct, no fluff. Structure: (1) thank them briefly, (2) recap the 3 most important decisions or alignments from the call, (3) list our next steps with owner and date, (4) list any open items we need from them with owner and date, (5) propose the next meeting. Maximum 200 words.

Meeting summary: [paste]

The two-meeting-rule of consulting: if your follow-up email goes out within an hour, you look on top of things; within 24 hours, you look adequate; later than that, you are losing trust. AI moves the median consultant comfortably into the first bracket.

Special-Case Prompts

Difficult Conversations

After a tense meeting:

Below is the transcript of a tense meeting where the client pushed back on our cost recommendation. Identify: (1) the underlying concerns the client expressed, beneath the surface objections, (2) what they may not have said out loud, (3) the 3 specific things we can address in the next email to rebuild trust, (4) draft the email.

Multi-Stakeholder Meetings

When 8 people are on a call:

This was an 8-person meeting. Build a stakeholder map: who supports our recommendation, who opposes it, who is neutral, and who has not spoken at all. For each, give the evidence from the transcript. Recommend who we should reach out to bilaterally before the next meeting.

Workshop Synthesis

After a full-day workshop:

Below is the transcript of a full-day strategy workshop with 14 client participants. Produce: (1) the 5 most important themes that emerged, (2) the moments of strong consensus and what was agreed, (3) the moments of disagreement and the underlying tension, (4) the 3 ideas that got lost in the room but deserve a second look, (5) the working artifact (a one-page synthesis we can email participants tomorrow morning).

Building an Institutional Memory

The meta-prompt that compounds across an entire consulting career:

Below are all the meeting summaries, decisions, and follow-ups from a 4-month engagement with a regional bank. I am about to start a similar engagement with another regional bank in a different geography. Extract: (1) the questions we should ask early that we wish we had asked earlier last time, (2) the decisions that turned out to matter most, (3) the assumptions we made that turned out to be wrong, (4) recommended adjustments to our discovery approach.

This kind of cross-engagement learning used to require a senior partner with a 20-year memory. AI makes it possible for a 3-year consultant to do the same.

Common Pitfalls

• Trusting the auto-summary. It is shallow. Always re-prompt with a structured request.
• Forgetting to disclose recording. This will end engagements.
• Skipping the "commitments" sections. These prevent the most damaging mistakes.
• Letting the tool send the follow-up automatically. Always review and personalize before sending.

Key Takeaways

• Use platform-native AI (Teams Recap, Zoom AI Companion) for client meetings; Otter or Granola for internal work.
• Always disclose recording verbally at the top of each call.
• Use a structured re-summarization prompt that captures decisions, commitments, and open questions — not just bullet points.
• Send the follow-up email within an hour. AI compresses this to a 30-second task.
• Periodically synthesize across many meetings to extract the evolving narrative — this is impossible to do manually.