The problem with asynchronous communication

In case you missed it, last week, Basecamp stirred up some controversy when they announced a number of changes, including that they're banning employees from discussing political and societal issues in the company chat. It turns out, this policy was inspired by a quasi-political discussion between one of the founders and several employees which turned ugly.

The internet has already discussed the merits of the policy change to death, so I'm going to leave that alone. Instead, I want to focus on the communication breakdown that led to the change, because I think there's something for all of us to learn.

When a conversation spirals out of control

If you read this article explaining what led to the drama, and the follow-up from one of the founders, you'll see that at first, everything seemed pretty reasonable. Back in the day, some employees did something that the company shouldn't have allowed (made a list of "funny" customer names). An employee brought up how problematic that is, and one of the founders apologized and shut the list down. So far, aside from the initial mistake of letting the list exist in the first place, this is how an internal issue like this should be handled.

But then, some employees kept discussing the issue in the company chat. We don't know what they said because only the founder's messages are public, but it seems like the conversation gradually got more and more combative until eventually the founders thought the best course of action was to pass sweeping new policies which caused about a third of their employees to quit on the same day.

While I don't have all the information, this fits a pattern that I've seen before, and I think there's a good chance this blow-up could have been avoided with more appropriate communication.

Asynchronous is deeply tied to remote work

Basecamp was one of the very early "remote-first" companies. They literally wrote a book called "REMOTE" which explained how companies can work without sharing an in-person presence.

A big part of most remote cultures is asynchronous communication. Rather than having most conversations in real time (meetings, phone calls, accidental conversations at the water cooler), the conversations move to asynchronous channels such as email, Slack, or even Basecamp itself (the product that Basecamp the company makes).

These async channels are great for a lot of types of productivity. With async communication, you can respond to conversations whenever you want rather than requiring everyone to be free at the same time. You can think deeply about what you want to say and take the time to edit the message. Most of the time, async communication is less distracting, clearer, and more concise than real-time communication.

But there are some problems with asynchronous communication:

Note: Yes, there are audio/video async channels such as Yak, Loom, etc. Those are a pretty minor part of the async ecosystem, so this post is specifically talking about text-based async communication like email, Slack, etc.

Async encourages mob mentality

Think about the most toxic online cultures you've ever seen. You're probably thinking of Facebook, Twitter, Reddit, etc. Sure, those are different from internal team communication at a business, but I don't think it's a coincidence that they're all text-based async channels. Why do async communities so often end being toxic?

Async strips away empathy.

Here are some specific ways this manifests:

  • Conversations are public to everyone by default. Instead of the 2-10 people that can reasonably participate in a face-to-face conversation, async conversations are normally open to everyone. It's much harder to emotionally connect with 50+than 10.
  • It's easy to instigate and bail on a conversation. Because conversations are public and people can comment even after the conversation has ended, it becomes much easier for instigators to fan the flames even after things have died down. Any random person at the company can pop into a thread and post a comment without necessarily thinking deeply about the impact it might have.
  • It's easy to misinterpret text. When you talk to people in person or via video, you pick up cues from their facial expressions, body language, and tone of their voice. With text, you read someone else's words in your own voice, and it's easy to misinterpret the tone they were trying for.
  • You care about people, but not pixels. Have you ever wondered why people get so angry when driving cars? I think it's at least partially that they stop associating the other cars with the people inside of them, and instead they become robotic combatants. That same thing happens when talking with people via just text. Even if you know them personally, they just don't seem as human.

The result is that public, async, text-based conversations are the perfect storm for reasonable people to end up in unreasonable disputes. Tempers flare, mobs form, opinions are interpreted in the least charitable way possible, and the shared humanity that holds us together weakens.

What should we do about this?

I want to be clear: I love async communication. I regularly push for more of it at LACRM. For the 99% of conversations where no one is at risk of getting mad at anyone else, I'm all about it. But as soon as a conversation starts to become combative, it should move to a more empathetic channel.

We actually have a bit of experience with this at LACRM. Back in 2017, a Google employee wrote an internal memo that leaked and caused some controversy related to diversity at tech companies. Like many tech companies, we had an internal discussion about the memo in our Slack workspace. It started to head in a direction where employees were taking sides and saying things that could hurt other people.

In that particular situation, the fix was almost stupidly simple: The few employees who were feeling particularly passionate about the discussion got in a conference room and talked face-to-face. Tensions dropped almost immediately. Everyone remembered that we're all friends. Whereas you might assume bad faith online, the second the conversation happened in person, we started assuming good faith again.

I'm not saying that real-time conversations will avoid all conflict, but it'll avoid a lot of it, especially the conflict that never needed to happen. With a real-time conversation:

  • The group is smaller, so things stay personal instead of tribal.
  • You're more likely to understand what other people are saying because of the benefit of tone, body language, etc.
  • You're more likely to give people the benefit of the doubt because you're reminded that they're a human and you care about them.
  • Once the conversation is over, it's over. It doesn't fester in a channel for anyone else to come fan the flames later.

Basecamp decided to ban all conversations that might spiral out of control (good luck with that btw).  If you're thinking about something similar at your company, I'd encourage you to instead try making a space for those conversations to exist in a less toxic way. It's a good thing for employees to challenge each other. It's only a problem when the conversation veers from respectful peers trying to learn to angry mobs trying to win an argument, and that's more likely to happen with asynchronous communication.

Update: Just a couple hours after I published this, some more information came out from the Basecamp drama. It turns out there was some fundamental conflict between employees that probably couldn't have been resolved strictly with a different communication medium, but I'd still argue that the rift between the founders and the ~20 people who resigned was probably avoidable if they'd communicated differently.

Have thoughts on this post? I'd love to hear from you! I'm @TylerMKing on Twitter.