Rebecca Hinds, head of the Work AI institute at Glean, says a good meeting is “a highly intentional meeting—intentionally scheduled, intentionally attended.” It leaves people feeling like they got a lot accomplished because they came together. It definitely could not have been an email.
The problem, she writes in Your Best Meeting Ever, is that most meetings fail to clear that bar. We spoke to Hinds, who will be speaking at Charter’s Leading with AI Summit Tuesday, about her new book, her recommendations for making meetings better, and the effect AI is having on meeting culture. (You can register to attend virtually here.) Here are excerpts from our conversation, edited for space and clarity:
In your book, you recommend conducting regular meeting resets. What’s the best way to conduct a ‘calendar cleanse?’
[When] I joined [Dropbox, a prior employer], as part of onboarding we learned about this ‘Armeetingeddon’ that had happened about a year earlier. A bot came in and deleted employees’ recurring meetings from their calendar. For two weeks, if employees had added recurring meetings back to their calendar, [the meetings] were auto-deleted at the end of the day. Ever since, I’ve been inspired by these resets and how you design them to be most effective.
Over the years, I’ve recognized the importance of designing this [process] so that employees have as much permission and autonomy to design their own calendars as possible. We know that activates the ‘IKEA effect.’ People value [things] more and they’re more likely to stick to [them]—whether it’s an IKEA desk or a newly rebuilt calendar—if they’ve had a hand in designing them. Encourage employees to collectively delete the meetings [themselves], then give them autonomy and advice in terms of how they rebuild their calendars.
You mention three types of meeting ‘guardrails’ in your book—speed bumps, gatekeepers, and blocks. How can leaders use them to reinforce the principles of a meeting cleanse after they’re over?
We need to raise the bar for what deserves to be a meeting. Speed bumps are the gentle nudges where you’re encouraging people to pause. For example, teams might use AI to have a pop-up [alert] that says, ‘Your meeting is now more than eight people. The research shows that it’s not likely to be as effective. Do you really need those eight people?’
The gatekeeper is a bit more intense. It’s an approval checkpoint—either a person or a technology has to sign off on the meeting before it’s scheduled.
Blocks are the most extreme. They’re hard stops. I was chatting with an organization a couple weeks ago [who has] used AI to determine whether a meeting is likely to be effective, and auto-delete it if it’s not. If you have a meeting where, 24 hours in advance, more than half the attendees haven’t accepted the invite—or you have an invite where there’s an expectation for an agenda and it’s not added—the AI auto-deletes it.
Ideally, blocks and gatekeepers aren’t needed and we can self-police and understand what a healthy meeting culture looks like. But in the meantime, having those nudges is a means of becoming more intentional.
How can teams measure the quality of their meetings?
My colleague Elise Keith first introduced me to this concept of ‘return on time investment’ or ROTI. Ask attendees to rate on a zero-to-five scale, ‘Was this meeting worth the time you invested?’ As a follow up, [ask] ‘what could I do as the meeting organizer that would allow you to boost your rating by one point?’
Pretty quickly, you’re able to identify those meetings where most of the participants are scoring zero to one. There’ll be a lot of them in most organizations. In many cases you’ll have split rankings, [with] half the attendees rating a zero to one, half rating a four or five. That’s probably an indication the meeting is relevant and useful for a specific group of people.
In general, I’m more concerned about AI being injected into our meetings than excited, but the thing I’m most excited about is being able to measure participation rates, [speaking] airtime rates, and engagement in meetings—never for the purpose of surveillance.
What makes you so concerned about AI in meetings?
We’re in more and more meetings where half the [participants] are bots and only half the room are people. People are more inclined to overschedule meetings because they can just send their bot. They’re more inclined to multitask in the meeting because they have this great transcription that they can refer back to after the fact. AI is amplifying a lot of the existing dysfunction associated with our meetings. It makes it easier for people to cognitively offload the really important human things that we know meetings are uniquely suited for.
If we are having more than one bot in a meeting, then it’s a sign that the meeting probably doesn’t need to exist. It’s probably not requiring that synchronous communication and it can probably be an email or an update in a project tracker.
Your Best Meeting Ever is out now. Order it on Amazon or Bookshop.
Read Charter’s guide to meeting resets here and our meeting facilitation toolkit here.
Register to virtually attend Tuesday’s Leading with AI Summit, where Hinds will be a speaker.