Why focus on AI meeting transcripts if the real problem is bad meeting prep?
meetrics.ai
meetrics.ai
And you can't ask people to submit questions or read documents or watch demo videos in advance, because they just won't do it. They'll say "my schedule is so full of meetings, I couldn't do the prep work". Well, it's full of meetings because we have 2x50 minute meetings to do what we should be able to do in one 25 minute meeting.
I can see you are well versed in corporate life.
Maybe just cut out meetings entirely and see how that goes.
> people have to read and formulate questions under time pressure
People who aren't good at that still have the option to read ahead of time. They need to know themselves to figure out which kind of person they are. They also have the option to submit feedback after the fact.
> Anyone who did read (or helped write) the documents in question is sitting around wasting their time.
Usually if you were an author or already read it, you either did your emails, wrote your next document, or just showed up after reading period. Sometimes you'd use that as your lunchtime if you WFH while you waited for people to read your doc. Or you would read the comments as they were being left on your document in real time so you could respond or do more research.
But I always managed to fill the time with something else productive. You adapt pretty quickly.
> And of course it doesn't make the meeting any shorter.
This is 100% true. It makes them longer and slows down decision making. But it also tends to mean that decisions can be made (and are recorded right in the doc!) in one meeting instead of multiple meetings, since the main thoughts are already written down.
To be clear, I'm not saying it's a good idea. I found most of the meetings run that way to be a waste, because they indexed on "every meeting starts with a document" even when it wasn't necessary. But it might be helpful to try it for one or two meetings where you are wasting a lot of time talking about things that could have been written down.
I'm making it such that if someone preps a meeting they HAVE to assign a responsible person for each and every talking point. When they do this, that person gets a notice saying they have something to prepare ahead of time for.
SO even if they don't have enough time to look at the whole agenda, they can easily identify what exactly they need to pinpoint-prep for.
I suspect this is why Amazon has adopted a policy of silent reading time at the start of a meeting. You can’t expect busy people to spend hours prepping.
Meanwhile, one policy may result in what you want actually happening, and another may not.
SO even if they don't have enough time to look at the whole agenda, they can easily identify what exactly they need to pinpoint-prep for.
1. Dedicate the first X minutes of the meeting to reading the document. It sucks, and you shouldn't have to, but clearly people aren't doing their homework, so you gotta give them time to do it during the meeting.
2. "I couldn't do the prep work" "Ah, I'm sorry to hear that. I'll cancel this meeting right now, since we cannot proceed without that; everyone can use the freed up time to go over the prep work required, and I'll re-schedule this meeting for later in the week."
The second option requires you to have both the confidence to say this, and the credit to unilaterally cancel-and-reschedule a meeting and to call people out in such a way. But if the meetings are useless anyway, cancelling and rescheduling one doesn't sound like a big loss.
Consider that the alternative in this scenario is not to solve the tasks in the meeting in the allotted time, but to waste time holding a meeting where things don't get resolved because people haven't prepared.
When they do this, that person gets a notice saying they have something to prepare ahead of time for.
SO even if they don't have enough time to look at the whole agenda, they can easily identify what exactly they need to pinpoint-prep for.
I noticed that I was in a ton of pointless meetings. Some were recorded, but I’d never go back and look at the recordings because the meetings were a waste of my time. I realized the common theme was none of these meetings were prepared well or at all.
One could look at meetings as a start-to-end process. AI transcripts and summaries are downstream of meetings. The problem is, if you have a bad meeting, you get bad AI transcripts and summaries - garbage in, garbage out.
Why not apply AI at the start of the meeting process where meetings are first created? Make people think of what they want to get out of the meeting, what they want to discuss, and who actually needs to be there.
In theory, this should result in people having:
> Fewer, shorter, and more efficient meetings > High quality agendas with worthwhile talking points > Healthy discussions that lead to an actual outcome
Therefore, the AI preparation tool should:
> Know what you want to get out of the meeting - i.e. what’s the objective? > Use AI to create an agenda with relevant talking points to hit said objective > Make you think of who should talk about each talking point and invite only those people > Bonus points if it follows a repeatable structure and does all of this very quickly, which should totally be possible with AI
What do you all think? Less AI transcription/summarization and more AI meeting prep?
I like the idea of declining a meeting if it's not prepared OR 'nudging' them to prepare ahead of time.
(so yes, I'd love anything that gatekeeps the creation of meetings in a way that ensures the meetings I get added to are actually worthwhile instead of having to work around it).
I don't think I've ever watched a recording of a meeting.
You mean something that now fits in an email?
If I can share anything or just talk through my experiences let me know. I think reducing time spent in nonsense meetings actually does make the world a better place for the humans in it, at least a little bit.
Let me know how we can chat - my email address is ashish.fernandez@meetrics.ai
Anyway, even with a software scribe, meetings should probably end with attendees reviewing and agreeing on some kind of written document which will be the meeting's artifact. You know, so people can refer to it later. Of course this never happens, because in meetings people get bored and tired and it's easier to just give everyone the Zoom wave and crawl to the sofa.
I found at uni that well-prepared blackboard/overhead lectures were excellent for taking notes. I caught many small typos and misunderstandings as I would actively work through each step of the theorem/proof/problem as I wrote.
In hindsight I realized they must have spent a great deal of effort in preparing the lectures precisely such that they were easy to write down.
In contrast, the PowerPoint lectures we had was exactly like you describe. Either I could focus on writing stuff down, but then I couldn't really follow what they were saying, or I could focus on what they were saying but not write anything down. Either way I felt I was missing out, as writing stuff down made stuff stick magnitudes better.
I attributed this to the fact that PowerPoint doesn't lend itself to be easily transferred to notes, so it requires a transcription step that detracts from the lessons.
I guess it's in large part due to a reduction in the number of secretaries and assistant roles in most companies, but while many part of those roles became redundant with word processing etc., producing good quality notes did not, exactly because meeting participants that need to focus on the meeting make really lousy note-takers.
Heck, I've been in meetings more than once where we referred back to the notes in the same meeting, and having good notes is amazing in those situations.
1. Even if you have a well run meeting, you should record it and transcribe it for searchability.
2. Ever worked with a third party? They probably suck at meetings, no matter how organized you are it won't matter.
- presenting information: is often better in document form (if precise, concise, not fluf, with diagrams etc.) with the added benefit of potentially being able to fix error retrospectively (Note: Similar to presentations you don't want to obsess about spelling, grammar, style and bullet points often are the way to go. Avoid "story writing" at all cost, you are presenting information not telling a story.
- technical discussions: if done in meeting for anything non trivial leads often to sub-par results, interestingly that is in my experience also true for fast messaging style text communication (like e.g. slack motivates you to do). Thinking more before writing and also having all you questions/comments written down before sending any of them does in my experience works the best.
- teaching people: here meeting like things can be good if individualized to the learner and supplementary only. But often you wouldn't call it a meeting anymore. Transcripts can be quite nice in this case but as it should only be supplementary should be unnecessary.
- synchronizing work: Inefficient if mostly done through meetings, through sometimes having supplementary meetings for it can be useful.
- planing work: Often highly inefficient if mainly done through meetings, through finalizing a plan through a short meeting is a good idea.
- brain storming: Can be started by a meeting, should not be concluded by a meeting as mind tend to pop up ideas at the strangest times and should not diverge into technical discussions about the proposals.
combine that with how interruptive meetings can be having too frequent meetings seems like a sure fire way to reduce productivity
I see this is an ad that promises "perfect meetings every single time". If you believe that, I got a bridge to sell you.
Maybe this product does make it easier to prepare good meetings. You will still have bad meetings, and you will still benefit from automated transcripts. Again, there's basically no reason NOT to have automated transcripts.
All of this data dumped in to Slack, so that it can easily be searched along with other conversations on the same subject.
The problem, is most people don't actually want to run a good meeting due to either having to accept the risk of whatever they're trying to do actually happening or not, or the risk of having to do other people's work if they run a meeting about it.
Corporations don't do things because they're better. They do things because somebody gets a promotion / comission / are easier / somebody bought some BS enterprise software.