Small steps towards getting rid of meetings
Meetings tend to suck, yet serve a purpose. How might we begin to eliminate them?
We all have them. We’ve all been in a lot of bad ones. And, depending on your job, you probably spend somewhere between 25-75% of your working life in them. I’m talking of course about meetings.
The problem with meetings
Meetings are one of the most common and popular ways of doing work today. Yet, they also tend to be one of the most inefficient, time consuming, and painful ways of doing work too. You probably don’t have to go very far back (maybe a couple hours at most) to a meeting that seemed like a waste of time, not just for you, but likely everyone involved.
If meetings can be so bad, why do we keep having them? Are there even alternatives? Or have we just been doing them for so long that we never stopped to ask the question why are we doing them in the first place?
There’s a reason meetings exist. Like all solutions, meetings attempt to address a problem (and sometimes more than one problem). Meetings address three main problems: gathering/spreading information, making decisions, and generating ideas. One meeting may try to address more than one of these problems at a time, but all meetings try to address one of these core problems. Whether or not they succeed at addressing these problems is the question of whether or not a meeting was worth everyone in its time. The best meetings set out with a goal of addressing one or multiple of these problems and then by the end of it, have actually addressed those problems.
In my opinion, one of the best ways you can improve the way an organization operates is by focusing on having better meetings. Really well run organizations have a higher % of good meetings than poorly run organizations.
The reason this matters is because meetings are very costly. Don’t believe me? Here’s an exercise you can do to see how much any one particular meeting costs your organization: take everyone in the room (whether physically in the room or otherwise) and take their annual salary, divide that salary by 52 weeks, divide that by 5 days, and divide that by 8 hours. This will give you a rough estimate of what each person’s hourly rate is. Once you have that number for everyone in the room, add them altogether. That’s how much a 1 hour meeting with those people costs. This can add up pretty fast, depending on the seniority level of people in the room and how many people are in the room. In fact, this little trick might be just what you need to convince people that it might not be worth it having a meeting in the first place, or ensuring only the people really required are in the room.
Shopify, for example, thinks this sort of thing might be so valuable that they have built it into their calendar system:
But straight up getting rid of meetings is not the answer either. Because, as mentioned above, meetings have a purpose, they are in theory supposed to address a problem (or multiple problems). Removing them entirely doesn’t necessarily mean those problems are going to be solved in other ways.
I wanted to explore this idea of how to get rid of meetings, yet still address the underlying problems that meetings are trying to solve. As a result, I’ve built two different potential solutions to two different types of problems. Both of these solutions are prototypes and as you will see, do not address nearly all the different types of meetings an organization might have. But, by building these prototypes and sharing them here, I wanted to give others the idea of what could be done to help their organizations start to eliminate unnecessary meetings, while still addressing the core problems that meetings try to tackle, which are again: information gathering and sharing, idea generation, and decision making.
An alternative to information sharing/gathering meetings
The most common of all types of meetings is the information sharing/gathering meeting. If you’re a manager, this is likely to be about 90% of your meetings. Yes, you make decisions but most of those decisions do not require a committee (and therefore a meeting) and you make most of them in the moment on your own. But what you actually need to make those decisions is information. In addition to gathering information, you also spend a lot of time sharing information. Maybe that’s in an All Hands type of meeting to share with a large audience or that’s in a smaller group setting to share information with other teams, etc. And even if you’re not a manager, you spend a lot of time in information meetings, like ones where you share and gather information with your teammates or gather/share information with larger groups within your organization.
As the most common type of meeting, there are many different forms and types of information gathering and sharing that could replace them. I choose to build a prototype for addressing one of the most common forms of these meetings that I have gotten to know over the years working in technology startups, which is that of the Stand Up meeting.
For anyone not familiar with it, Stand Up meetings are a pure information sharing and gathering meeting. Everyone on the team “stands up” (although, typically not literally) and shares what they worked on since the prior meeting, what they are working on today, and if they have anything getting in their way from getting something done. These meetings can be valuable because they can help uncover issues and address them quickly, course correct things that may be taking more effort than they are worth, or make sure that everyone has a clear understanding about what they should be doing on any particular day. These meetings can take both a synchronous or asynchronous form. Usually if asynchronous, it’s done digitally and synchronous, either in-person or over video chat.
While that sounds great in theory, in practice, they almost never are like this platonic ideal. Instead, what typically happens is: 1) it’s hard to coordinate everyone to meet at a given period of time, especially with varying time zones, time off, etc, therefore most of these move to an asynchronous form, however that causes 2) not everyone provides timely or accurate updates, since they are busy, forget things, etc, which of course then causes 3) people don’t engage with each others’ updates, which lead to 4) worse updates over time, effectively making the whole purpose of the meeting less and less valuable over time. In the end, you just end up reaching out to everyone individually that you need to, which isn’t the most efficient use of everyone’s time and can disrupt people’s work.
The ideal solution for this then would be to have everything typically done in a standup meeting done automatically, making it super convenient for all to participate, while still getting the information to everyone who needs it.
So that’s what I was trying to do with Team Digester. Team Digester is a simple Slack bot and Github integration that automatically shares team-wide updates with everyone based on what everyone on the team has been up to. The way it works is that it integrates with Slack to provide the messages on a periodic basis, which can be customized per user, so that you get the update at the exact time you want it, not having to rely on others to share at that time. These messages are then based on the activity from your organization’s code repositories on Github.
Since I was targeting the engineering team Stand Up meeting, updates on Github was the first thing I wanted to demonstrate, although you could update it to work with many other types of integrations and activity. If you were a marketing team, for example, and used Asana or Trello to track the activities of your team, you could modify the app to use those integrations as the source of the updates. You could even integrate it with a company calendar, like Google Workspace, to use everyone’s calendar for tracking what they are doing.
Below are some of the screenshots for what it looks like:
Note: it’s pretty limited right now in functionality and really can only provide basic updates for your team. Although this is still better than the alternative of not getting updates at all, or getting poor ones, it still may leave a lot to be desired.
Since this was an experiment and there’s a lot to be improved with it, I’m making it completely free and open source. You can therefore use it with your teams however you want. There’s a bit of setup involved in running it on your own which you can view here, but if you would like an entirely hosted version, just let me know. If enough people actually want something like this, I’ll see what I can do to put a hosted version together.
An alternative to idea generation meetings
The other category of meeting that I tried creating something to replace it with was the idea generation meeting1. This category of meeting is likely to be the least common meeting form that you will have. It depends on your role at a company, however, and some of the more creative types of jobs might do these more often than others but, even so, I’d bet that they’re not spending more than 10% of their time in these types of meetings.
However, in information businesses, like software businesses, ideas are the currency in which products and businesses are made with. Take the other two types of meetings, information gathering/sharing and decision making. Information gathering/sharing meetings are meant to help generate ideas, while decision making meetings are meant to then decide which ones to act on. Therefore, idea generation meetings are sort of like the glue that binds the other two more common meeting types together. So, despite idea generation meetings being infrequent, the output of those meetings should be important.
The question then is why don’t you have more idea generation meetings, if ideas are so important? I think there are two main reasons that idea generation meetings don’t happen that often: 1) they tend to not generate many valuable ideas and 2) they seem to be the most inefficient of all meetings to the participants. These two things are related, let me explain each in slightly more detail.
Idea generation meetings don’t tend to generate many valuable ideas because most organizations do them wrong. Great ideas don’t solely come from people sitting in a room for an hour throwing out the first thing that comes to their mind and writing them all on a board and then voting on the best ones. Ideas take time to develop in our minds. They take periods of quiet thinking, alone. Then once you’ve given it some time and put in the effort to really think about things, it can be valuable to bounce those ideas off others and build off of other’s ideas2. But that’s not how these meetings are typically run.
Which leads us to the other reason they don’t happen often, because they feel inefficient. They feel inefficient because instead of having everyone take the time outside of the meeting to actually spend thinking, these meetings typically ask people to do that thinking in the meeting itself. But given the cost of meetings that we just talked about and how hard it can be to coordinate people for them, the meeting’s length is likely not going to be very long, so at most you dedicate 15 mins towards actually thinking and generating ideas.
Maybe you can think of great ideas in 15 mins but I sure as hell can’t. It also doesn’t help that sitting in a meeting with others silently for 15 mins doesn’t feel productive. And while it would be fine to not feel productive if you got good outcomes, spending only 15 mins on ideas, means that the ideas generated are likely to be bad. It’s no wonder why these meetings aren’t held often and don’t end up with ideas you use.
But the problem with not holding these meetings is that I think companies under-generate new ideas because they never spend the time to actually think about them. Many organizations run by the calendar. If it’s not on the calendar, it’s not being addressed. That’s all fine and good for the latest productivity hacks from Twitter gurus but it can result in never making the time to think or explore genuinely new ideas.
With this in mind, I started tinkering with the idea of automated idea generation. One of the great things about the latest AI models is that they tend to be really good at producing output, no matter what. For a lot of things, this can be bad because the AI will generate something and therefore generate made up stuff if it doesn’t have the perfect answer. However, this bug is actually a great feature for idea generation, since ideas are always “made up”. In fact, the more novel an idea is, the more chance it has in being something of a lot of value. So, if you ask the AI for a set of ideas, it’s going to come up with a set of ideas, without fail and fast. It also does this quite cheaply.
Now, I don’t think that the current AI will necessarily be better at generating a singular good or novel idea than a human would, but sometimes creativity and ideas are not about quality but more quantity. The more ideas you have, the more likely it is you’ll be able to spark even more ideas or combine them into new ideas. In fact, for some creative professions like writers, there are specific exercises meant to do exactly this, generate a list of places, characters, and professions and try combining them in new and different ways to spark ideas.
That all said, it is also the case that AI actually does pretty well on quality too. There are some new papers showing that AI not only produces more ideas than humans, but also tends to outperform them on some measures of quality too. This then presents us with a great alternative to idea generation meetings: using AI to generate ideas automatically.
And that’s exactly what I was trying to do with this program that I built. It’s a short program that’s purpose is to help you generate a list of ideas that you can then use for whatever purposes you want. If you want to evaluate the list based on some internal measurement or even with your customers, you can do that. And just like the other application I shared above, this one is also completely free and open source.
The easiest way to use it is to click this link and then copy the notebook into your own Google Drive account. Then you can just follow the notebook’s instructions. If you don’t use Google, you can find all the code to run yourself in this code repository on Github.
You could of course just use something like ChatGPT to do this but one of the things that I wanted to also do was provide the output in a form that you could then use to read through and evaluate the ideas with. In addition, making it a program means that you can update it to run automatically, periodically, with any input your want, without having to copy and paste things over and over.
The program does this by allowing you to quickly generate a CSV of 10, 20, 30, etc ideas for any particular subject matter. You can then use that list to do your own idea evaluation (for example, you could measure them on: feasibility, effort, opportunity, competition, etc).
It’s a simple script and not meant to replace actually generating some ideas yourself, but the magic of it is that if you’re ever stuck on something, you can just run the script and get 100s of ideas in minutes, for less than the cost of a coffee. No need to call a meeting. And even if you want to call that meeting, everyone can go into it with a list of their favorite ideas, generated automatically by AI. No more excuse for anyone to not show up with some ideas.
Conclusion
Everyone has meetings, few really enjoy them, and most are a waste. We need to start working better, smarter, and more productively. The application and script that I shared in this post are hardly going to change the way anyone works today but I do think they give a glimpse of how we might in the future.
If nothing else, I hope that this post makes you start to question the next meeting you throw on the calendar and whether or not it can be replaced with something else.
While I didn’t build anything to help replace the last category of meetings, the decision making one, I have built a number of other tools around decision making generally. At some point, I might be interested in tackling this meeting form as well with an alternative solution but for now that’s the only one I don’t have a particular tool for.
If you want to learn more about how to generate great ideas, read this essay on creativity by Isaac Asimov.