For many people today, most work happens in information environments. Much of it consists of collaborating on and coordinating activities. In other words, it requires communicating with other people. There are various ways for people to communicate in information environments. Email is one of the oldest. It’s also one of the best.

You can set communications channels on a continuum based on latency. On one end of this continuum, you have face-to-face communication, which is very low-latency. When somebody says something to you in your presence, you get what they’re saying almost as soon as the words leave their mouth. In many cases, there’s also a social expectation that you will reply right then and there. Long pauses can be awkward; you must respond even if it is to say, “I need a minute.” Face-to-face oral communication is what we call a synchronous channel — the back-and-forth between participants happens in “real time.”

On the other end of the continuum are communications channels that don’t carry such expectations, mostly for technical reasons. For example, if you send a question on a (physical) postcard to your friend halfway around the world, you know that the piece of cardboard that conveys your question will take some time to get to where your friend is. Your friend might read your message a few days from now. If she chooses to respond with a postcard of her own, it too will take several days to reach you. Considerable time will elapse between the time when you issued your question and when you got a reply. Thus, postal mail is what we call an asynchronous communication channel.

When you’re interacting over asynchronous channels, you don’t expect to receive an immediate reply. You issue your statement or question, and let it go with the expectation that sometime down the line you may (or may not) receive a response. Of course, this won’t work for many communications tasks. For example, you wouldn’t want an asynchronous emergency response system. And if you’re making breakfast and ask your spouse what he or she wants to eat, getting their response in the evening won’t do.

Synchronous communications channels are good at coordinating these types of activities, which require quick responses. That said, not all do. And asynchronous communications carry a high cost: they require the sustained attention of both parties. For an emergency response system to work, somebody must be on standby, ready to pick up the phone the instant you call. And when they do answer your call, you’d better be prepared to provide them the information they need, right then and there. So both of you must be focused on the conversation.

In our continuum, email sits closer to postal mail than other communications channels. It’s an asynchronous channel. Right now I have an email in my inbox that was sent by somebody in Israel. This person sent the message while I was sleeping last night, and I’m sure he wasn’t expecting an immediate reply. This is in part because he probably knows I’d be sleeping when the email landed in my inbox. But it’s also because we understand this is how email works. We bring expectations to the channel, much as we do when we’re communicating face-to-face.

As a result, email is not good for dealing with urgent requests. But it’s a worthwhile tradeoff, since much of the stuff we communicate (especially at work) isn’t urgent. And email also allows us to keep control of our attention, so we can be productive in other ways “in between” communications. Instead of being in constant standby waiting for a message, email invites you to process it at your own pace. For many cases, replying to an email after a day or two is no big deal, and some emails require no response at all.

With the rise of instant messaging over the past few years, this equation has changed. IM is closer to the face-to-face end of the continuum. (That’s why it has “instant” in its name.) When you send a text message to someone, you expect a reply back fast. IM isn’t as fast as talking with somebody face-to-face — after all, you still have to type messages — but it’s still faster than email.

The tradeoff with IM (and systems modeled on it, such as Slack) is speed for quantity and length. Structurally, IM encourages you to write shorter messages. You send one request with the expectation of receiving one reply. Because messages tend to be shorter and more casual, they’re easier to dash off. You end up generating more shorter messages rather than fewer longer ones.

Synchronous channels are enticing. Getting a quick answer can help you move faster when you’re trying to get something done. Thus, we may be tempted to move our communications to channels that encourage (close to) real-time response. But again, not everything requires immediate attention; moving discussions to channels that encourage quick responses tends to make everything — even trivial matters — urgent.

A couple of years ago the team I was working with started using Slack for our internal communications. Little by little it came to replace email for much of our work. Our messages became shorter, which was good. However, soon there were more of them than when we were using email. It became​ easier to message everyone in the team, so we did it more often. Messages were also less focused: we started mixing in more jokes and random observations along with task-relevant stuff. While this made work more fun, it also lowered our signal-to-noise ratio, impacting our productivity.

Another effect of this move: recency trumped relevance. If I started a thread about project x in the morning, and somebody injected an aside about project y, the conversation could veer off in that direction. A colleague joining the chat room in the afternoon may completely miss the discussion about project x. (I don’t use the word “chat” casually here: our work discussions became more like chit-chats than considered discussions.)

Finding stuff later also became more challenging. Where was it that we noted that site’s URL? What did we decide about project X? Questions such as these became surprisingly difficult to answer since the relevant messages were now buried among hundreds of other tiny morsels. Yes, it’s possible to search in IM and Slack, but search in these systems tends to be worse than in email. And when you do find stuff, much of the context around it is lost. Information becomes more granular and less meaningful.

Contrast this with email: Searching for information in most modern email systems such as Gmail, Outlook, and Apple Mail often returns what you looked for in the context of the discussion thread where it happened. Where I sometimes had a hard time finding months-old information in Slack, I have an easily searchable email archive going back over fifteen years. It’s an invaluable resource that has saved my skin multiple time. (This is another benefit of email: its longevity means more mature tools for accesing and managing it. But that’s a story for another post.)

Synchronous communications demand your attention and discourage contemplation. The emphasis is on responsiveness, not thoughtfulness. Writing an email carries a higher cognitive cost than sending an instant message. When you write an email, you must consider who the recipients are, what the subject is about; you must think about what you’re doing. In other words, you must write. You don’t “write” an IM in the same way; you shoot it off. I suspect this explains the popularity of IM; many people don’t like to write. They just want to get on with whatever they’re doing.

But when what they’re doing is collaborating in information environments, being more considered can be a good thing. Writing an email slows you down. And when you read an email, you have time to think about the response. This slower frequency of communication can reduce the amount of noise in the system, allowing team members to have greater control over their attention. While this may initially result in the team moving a little slower, ultimately it will lead to a better use of everyone’s cognitive resources.