Three ideas that landed
First, most meetings are not badly run. They're badly designed. We try to facilitate our way out of meetings that shouldn't exist in their current form. Rebecca's reframing was precise: treat meetings like products, not habits. If there is no clear user need, owner, and outcome, you're shipping a broken product on repeat.
Second, discussion is not a decision. Many leadership teams are excellent at conversation and then surprised by slow execution. If the goal is a decision, design it. Name the decision. Name the decision owner. Define what done means. Capture it in a way that survives the meeting.
Third, recurring meetings are a silent budget line. They often become the largest ungoverned spend of leadership capacity. They started for a reason, but the business changes and the meeting doesn't. If you don't periodically re-justify recurring meetings, they become institutional drag.
The AI angle
This connects directly to what we teach. Non-technical leaders do not need more AI inspiration. You need AI leverage.Used well, AI improves meeting quality in practical ways: turning messy inputs into crisp pre-reads and agendas, making decision points explicit, capturing actions and owners cleanly, creating follow-through that doesn't depend on one heroic person, and noticing participation patterns that would otherwise stay invisible.Not AI for AI's sake. AI for clarity, focus, and execution.
Your 14-day challenge
Pick one recurring meeting. The one with the biggest attendance list, the least clarity, or the most "we always do this." Then run this redesign:
- What outcome must it produce?
- Who owns that outcome?
- What can move async?
- What decision is required, and how will it be made?
- What will we stop doing?
Reply with the meeting name and the first change you'll make. We'll share the best redesigns back with the community.
We started the Masterclass with meeting-bot whack-a-mole. People were live-reporting the calendar invites they were about to delete, decline, or redesign. Therapeutic doesn't quite cover it.
More soon,
.webp)
