Why email quietly wins for distributed teams (and chat quietly loses)
Real-time chat feels faster, but the research on interruption cost and email stress says the opposite for a specific kind of work: outreach and follow-up.
By David Lara, Founder
Founder-reviewed ·How we research and correct articles
Slack feels faster than email. That feeling is real — a chat message does get read sooner, on average, than an email sitting in a queue. What that feeling obscures is a different question entirely: faster for whom, and at what cost to the work actually getting done well. For a specific, common category of work — outreach, follow-up, anything that depends on a considered reply rather than a reflexive one — the research on interruption and attention says email’s slowness isn’t a bug real distributed teams should be racing to fix. It’s closer to the point.
What a chat notification actually costs
The clearest evidence here isn’t about email or Slack specifically — it’s about interruption, and it predates modern team chat entirely. Gloria Mark, Daniel Gudith, and Ulrich Klocke’s CHI 2008 field study observed information workers across normal days and days with interruptions removed, and found something counterintuitive: interrupted workers didn’t get less done. They compensated by working faster to make up the lost time — and paid for it in measurably higher stress and frustration. Speed was preserved. Something else wasn’t.
That’s the honest tradeoff a constantly-open chat channel makes on a team’s behalf, dozens of times a day, without anyone voting on it: every notification either gets answered immediately (interrupting whatever was actually being worked on) or ignored (defeating the reason it exists as a “fast” channel in the first place). There’s no third option where real-time chat delivers speed without a compensating cost somewhere.
The email-specific version of the same finding
A separate, more directly relevant study makes the mirror-image point about email itself. Kushlev and Dunn’s two-week field experiment, published in Computers in Human Behavior, had 124 adults limit email checks to three times a day for one week and check unrestricted for a control week. The batched week produced measurably lower daily stress. The finding isn’t “email is calming and chat is stressful” in some inherent sense — it’s that batched, intentional checking of a channel beats constant, reactive monitoring of it, regardless of which tool sits underneath. Email just happens to be the channel that’s culturally acceptable to batch. Chat, by convention, mostly isn’t — leaving a Slack DM unread for three hours reads as ignoring someone in a way leaving an email unread for three hours doesn’t.
Where each one actually fits
None of this makes chat useless — it’s the right tool for genuinely time-sensitive coordination between people who are online at the same moment. The mismatch shows up specifically when chat’s immediacy convention gets applied to work that doesn’t benefit from immediacy: a follow-up that deserves a considered reply, an outreach sequence spread deliberately over days, a reply that needs research before it’s worth sending.
| Situation | Chat fits better | Email fits better |
|---|---|---|
| Time sensitivity | Genuinely blocking, needs an answer now | No real reason it can't wait a few hours |
| Reply quality needed | A quick yes/no or status check | A considered response that benefits from a pause |
| Audience | Someone online right now, same team | External contact, unknown availability or timezone |
| Cadence | One-off, resolved and done | A sequence spread deliberately across days |
| Documentation value | Low — meant to be ephemeral | High — a durable, searchable record |
Cold outreach and its follow-up sequence sit almost entirely on the email side of that table by design, not by accident: the whole premise of a good sequence is that the second and third touch land days apart, giving a prospect room to actually consider the first one rather than reflexively dismissing an unexpected interruption. The reply-rate curve across a follow-up sequence depends on that spacing working as designed — a mechanism that a chat-speed, always-on channel structurally can’t replicate for someone who isn’t already expecting to hear from you.
What this means for how a distributed team should actually work
The practical takeaway isn’t “ban Slack.” It’s narrower: protect the batching convention that makes email work, and stop treating every channel as though it needs the same instant-response norm. Plain-text, unhurried email already reads as more genuine to a recipient than something visibly optimized for speed — which tracks with everything above. The channel that lets both the sender and the recipient think before responding tends to produce work that holds up better than the channel optimized purely for how fast a reply can arrive.
Built for the channel that actually fits this work
Norbelys is built around exactly this fit: sequences that space follow-ups deliberately over days instead of firing reactively, and a unibox that gathers every reply into one durable, searchable place so a considered response doesn’t get lost the way a Slack thread three days old tends to. If your outreach workflow currently lives in a scramble of ad hoc messages and manual follow-up reminders, that’s the chat-speed failure mode showing up in a channel that was never meant to run that way. See Norbelys’s plans and start running outreach on the channel the research actually supports.