Office culture is religion, but there is also data
About a zillion words have been written on the internet over the last few years about the relative merits of working in person vs. working remotely. You believe what you believe, and I am not going to change your mind about it with this post.
People have religion about this topic, and for all kinds of reasons. Advocates on both sides argue why their preferred arrangement is clearly better for productivity and/or culture. It's kind of a judgy conversation on both sides, if I'm being honest.
Luckily for me, I am not running a company in 2024, so I can just nerd out with the data rather than getting evangelical!
Earlier this year, I began collecting recorded workplace meetings from as many different teams as I could. I now have over 1,100 hours of meeting data from more than 150 different teams. So let's find out: Are in-person or remote meetings more inclusive?
Let's nerd out with the data
For this analysis, I included 200 hours of meeting data: 100 hours of meetings that were fully co-located (i.e. everyone in the same physical conference room), and 100 hours that were fully distributed (i.e. everyone dialing in via video). All meetings had between 2-10 attendees.
I analyzed the meeting data to answer three questions:
- How does the number of people in the meeting change participation patterns?
- Did attendees' race or gender have an impact on participation patterns?
- Did the in-office vs. remote meeting setting make a difference to either of the above? Is one setting more inclusive than the other?
Forget hot takes, let's measure
Fans of remote work often argue that it is more inclusive. When you hire without regard to residential location, you open up new talent pools for consideration. When people aren't bound by an in-office schedule, they can get their work done more flexibly, meaning that you can employ more kinds of people with varied life circumstances.
That may all be true as far as hiring diversity is concerned, but the meeting data tells a different story about inclusion.
For each meeting, I did a very simple analysis: I counted up the number of attendees who verbally participated in the meeting. I counted only substantiative contributions; just offering greetings or pleasantries did not count. In my analysis, a meeting where every attendee contributes would be the most inclusive, and a meeting where only a small fraction of attendees contribute would be the least inclusive.
The bigger the meeting, the fewer people participate
The more people in the meeting, the lower the percentage of attendees who contribute. This is not surprising and it is true for both in-person and remote settings. For instance, in all the meetings with only two attendees, both attendees spoke, making a 100% attendee participation rate. But in meetings with 10 attendees, it was common that half or more of attendees didn't speak at all once the meeting began.
No matter how big the meeting, the average participation rate is higher in person
By the time an in-person meeting has 9-10 attendees, only 55% of people are contributing. But when the 9-10 person meeting is remote, the participation rate is significantly lower, with only a third of people contributing.
The difference in participation rate between in-person and remote meetings increases as meeting sizes get bigger. When meetings have just 2-3 attendees, participation rates look quite similar in both office and remote settings. But by the time a meeting has 6+ attendees, remote meetings have an additional 25% of attendees not participating at all compared to in-person meetings.
The bottom line: Remote workplaces may have an easier time hiring a broad range of people, but the remote meeting setting just isn't as good at including everyone in discussions.
Remote meetings are even worse for women's participation rates
As we just saw, participation rates go down across the board when meetings get bigger. They go down even more in remote settings. I also wanted to know whether office or remote settings worked better for underrepresented groups.
I have data about participant race and gender, so I looked at both dimensions. I previously used the same broad data set to explore idea annexation, the phenomenon where someone gets credit for making a comment that was previously made by someone else and ignored. I found that women are significantly more likely to have their ideas annexed than men are), so I was expecting to find some impact in this case too.
Race didn't make a difference to meeting participation rates in this data, but gender sure did. In both office and remote meeting settings, women's participation rates decline faster than men's as meetings get bigger. But the remote setting has a particularly chilling effect on women's participation rates.
When an in-person meeting has five attendees, 4% fewer women than men participate in conversation. When the meeting is remote, 9% fewer women participate compared to men.
As the meeting size grows, the gap between women's participation in office vs. remote settings widens quickly. By the time in-person meetings have 10 attendees, 12% fewer women participate compared to men. But in remote settings, 30% fewer women than men participate.
The bottom line: While remote meetings drive down participation rates for everyone compared to how they'd show up in person, the effect is particularly staggering for women.
What do you think?
Thanks for reading!
Kieran
Catch up on nerd processor case studies | Subscribe | nerdprocessor.com