Meetings

Meeting tooling has changed more in the last three years than in the previous fifteen, mostly because transcription got good enough to be assumed. The launches here cover recording and transcription, automatic notes and action extraction, agenda and facilitation tools, asynchronous alternatives, and the analytics layer that tells an organisation how much of its week it is spending in rooms.

The transcription-and-summary group is the largest and the differences are practical rather than technical. Nearly all of them transcribe accurately enough now. What separates them is speaker attribution in a messy conversation, handling of jargon and names specific to your work, and whether the summary captures decisions and commitments rather than producing a fluent paragraph about what was discussed. A summary that says "the team discussed pricing" is worthless; one that says who agreed to do what by when is the entire value.

The consent question is not optional and is frequently glossed over. Recording rules differ by jurisdiction, some requiring all parties to agree. Tools that announce themselves in the room and make the recording state obvious are protecting you as well as the other participants. Tools that join silently are creating a problem you will own.

The asynchronous group is the more interesting one strategically, because the highest-value meeting tool is often the one that prevents the meeting. Structured written updates, recorded walkthroughs with threaded replies, and decision documents with an explicit deadline for objections replace a meaningful share of recurring meetings outright, and they leave a searchable record by default.

Facilitation tools — timed agendas, round-robin input, silent-writing phases, anonymous input before discussion — reliably improve the meetings that do need to happen. They work by removing the structural advantage of whoever speaks first and loudest, which is a real effect and larger than most people expect.

For the analytics layer, the honest caveat is that measuring meeting load changes behaviour more than any recommendation the tool makes. Simply publishing the number of hours a team spends in recurring meetings tends to cause some of them to be cancelled.