Abstract:

The article explains why many people sleep poorly even after closing their laptop—not because of screen time, but because modern work creates a “coordination-ready” brain stuck in warm standby, quietly monitoring for the next Slack ping when norms around response time and urgency are ambiguous. It argues this fatigue comes from coordination debt (dependency work with no clear “done” signal) amplified by fragmented tool stacks, cross-posting, and “just checking” follow-ups that make urgency get inferred from noise, so the mind keeps running background reputation math and bedtime simulations (“did they see the doc; am I blocking the release?”). Instead of adding another sleep-hygiene checklist, it proposes upstream communication design: a minimal “response contract” by channel (with clear reply windows), one explicit urgent lane defined by impact × urgency (e.g., outage, security risk, data loss, hard deadline) to prevent multi-channel escalation, and a separation between an acknowledgment (“seen + next update by X”) and real work to avoid fast-reply theater. It also recommends small message metadata that closes open loops (“need approval by 17:00… otherwise I ship option B”), a 3-line end-of-day broadcast (Changed/Next/Blocked + owner) as a cheap status page your brain can trust, and quiet-hours norms where comments are allowed but replies wait until morning, supported by simple scripts and a 7-day pilot measured by whether you wake up wanting to scan messages and whether sleep feels continuous or like standby—while noting persistent insomnia or signs of sleep apnea warrant clinical evaluation and CBT-I.

The laptop shuts. Slack goes quiet. You still don’t fully go offline. I’m in Lisbon now, still at the desk, often past midnight, and sleep is the variable I have not solved.

You get 7 or 8 hours, but it feels light and easy to interrupt. Like your brain is still half working. Not because you love screens. More because it feels like response time is being watched, even when nobody says it. So you half-sleep and keep a small part of your attention “available” for the next ping, the next question, the next vague “quick check.” I’ve done this on bad chairs and worse desks across Europe, thinking it was fine because nothing “hurt.”

This article is for that specific kind of tired. The one that comes from coordination, not pure workload. Taskwork can end. Dependency work often doesn’t, because there is no clear done signal. And when norms are fuzzy, the brain fills the gap with monitoring and rumination. Did they see the doc. Am I blocking the release. If I don’t answer now, will this look careless.

What follows is not another bedtime checklist with a new rule to fail at. The better lever is upstream communication design that makes offline time predictable and boring. You’ll see how coordination debt shows up in modern tool stacks, why urgency gets inferred from noise, and how a small response contract by channel can make silence normal again. Then it gets practical with simple message details that close open loops, a tiny end-of-day update that replaces late-night checking, quiet-hours norms that don’t turn into fast-reply theater, and a 7-day pilot approach that fits inside an overloaded week.

The coordination-ready brain

Sleep that never fully powers down

The room is dark, but your brain is still in “quick check” mode. You get 7 or 8 hours, but it feels light and interruptible, like you could switch into checking in 10 seconds.

This is not always about screens. It is often about the sense that response time is being noticed. Research calls this low psychological detachment (you don’t mentally leave work) and telepressure (the feeling you should reply quickly, even when nobody asked you to).

Warm standby is not rest

Coordination-ready means you are not doing the work, but you still feel on call to align, reassure, and triage. The loops are rarely “how do I solve this” and more “did they see the doc” or “what if I’m blocking the release and don’t know it.”

That is rumination. It keeps the mind loud right when it should slow down. Think of a service kept warm because traffic might show up and a cold start would look bad.

Ambiguity fuels vigilance

Unclear norms create uncertainty about what is “late” or “bad.” Uncertainty feels risky, so the brain compensates by monitoring. Even without notifications, it keeps polling.

The internal dialogue is boring but sticky:
“If I don’t answer now, will this look careless?”

This is upstream, not a bedtime project

This is not another sleep hygiene checklist with a new rule to fail at. The better lever is upstream communication design that makes offline time predictable and boring.

Norms, not willpower, set expectations. Left to myself, I will work through the night, so I don’t trust willpower here. If the system keeps suggesting “anything might happen,” sleep will stay in standby.

Coordination debt beats workload

Taskwork ends, dependency work does not

Execution load is producing the thing. Coordination load is aligning, clarifying, updating, waiting, preventing misunderstandings.

Execution can end. Coordination often does not, because it has no clean done signal. So it leaks into evenings as “just one more check.”

Tool stacks multiply reassurance work

Slack plus email plus tickets plus docs turns a normal contributor into glue. People hedge with redundancy because visibility matters.

• cross-posting the same request
• follow-ups that start with “just checking”
• tiny acknowledgments to prove responsiveness

Fragmentation and interruptions increase stress in real workplaces. After-hours reachability erodes boundaries and recovery.

Urgency gets inferred from noise

When nothing encodes priority, urgency gets inferred from social signals. A message lands in Slack, then “for visibility” via email, then again in a ticket.

Even if nobody meant escalation, duplication looks like escalation. Over time the brain learns a bad rule.

Night pings might be the start of something messy, so stay alert.

A quick recognition check

Non-medical self-check:

• waking with an urge to scan status
• sleep feeling like standby
• mentally simulating who might be blocked by your silence

That is detachment failing in a very specific way. The fix starts by making response expectations explicit and boring.

Response contracts that let silence be normal

A minimal response contract

A response contract is not a lifestyle poster about “boundaries.” It is removing ambiguity so the brain can stop doing background reputation math.

Think of it like a clear “when to expect an answer” rule. When it’s clear, people stop checking. When expectations are vague, telepressure fills the gap.

Keep the contract tiny and easy to remember:

• Chat during business hours. Respond in 2–4 hours. Response means ack plus an ETA if needed
• Tickets. First response by next business day. Updates go in the ticket as source of truth
• Email. 24–48 hours, not a same-day channel
• After hours. No response expected by default
• If you need faster, use the urgent lane. Do not multi-channel it

Urgent needs 1 lane, otherwise everything becomes urgent by social inference. A simple definition is impact × urgency. Urgent is actionable and time-critical, like a customer outage, security risk, data loss window, or a hard external deadline.

To avoid a new sport called fast-reply theater, separate response levels:

• ack. “seen, next update by X”
• meaningful reply. decision, answer, or next step
• resolution. done or moved to a tracked plan

Make the default offline safe

Norms fail in busy weeks because they rely on memory and good intentions, which are not scheduling tools. Defaults and templates are more reliable.

Treat it like a small team rule change, not a culture manifesto:

• pick 2 rules
• publish them where people actually look
• keep surface area small

Avoid framing that accidentally normalizes the bad baseline. “Stop messaging after hours” implies everyone does it. Better is “After hours, no reply expected. If it’s urgent, use the urgent lane.”

Clear messages that stop bedtime simulations

The stimulant is often the payload

The screen is not always the stimulant. The message itself can be.

“Quick question” or “Got 1 minute” creates an open loop. The brain keeps it in memory until a plan exists. Then it turns into more mental simulations right when you want quiet.

Minimum viable request metadata

Role ambiguity turns silence into risk, especially for the glue person who gets pinged in 3 channels and triages anyway.

A practical fix is slightly better structure in the first message. Minimum structure, not ceremony:

• what is needed
• by when
• how urgent
• what happens if not

Before after in 1 line:

“Quick question” becomes “Need approval on X by 17:00, low urgency, otherwise I ship option B.”

Acknowledgment that reduces uncertainty

Ack-only helps when it reduces receipt uncertainty, not when it creates a new mystery.

Bad ack:
“Seen.”

Good ack:
“Got it, no action tonight, next update tomorrow 10:30.”

A tiny end of day update that replaces checking

A 3 line broadcast beats 12 follow ups

A short async update before a cutoff reduces dependency guessing right before bed. Many interruptions are just people trying to find the current state.

Use this format:

• Changed
• Next
• Blocked plus owner

Why sleep cares about this

It removes “what is the current state” uncertainty. That makes detachment easier and reduces the itch to check. It’s like a cheap status page your brain can trust.

Guardrails:

• short
• linked to real work artifacts
• used by someone

Optional, low-ceremony check:
count after-hours “any update” pings per week before and after.

Quiet hours that keep work closed

Late-night comments are “optional” on paper, but they still land like a small request when rules are fuzzy. When “done” can be reversed at any moment, the evening feels provisional.

A quiet-hours norm keeps feedback moving without making response a midnight sport:

• comments welcome after cutoff
• replies happen tomorrow
• risks stay safe to raise

Template line:
“No reply needed until tomorrow. If you see a risk, please flag it.”

To reduce social guessing games, define a tiny reopen ladder tied to impact × urgency:

• compliance or security risk
• customer impact that cannot wait
• hard deadline that will fail overnight
• data integrity risk or loss window

Add deferral logic so people do not escalate just to feel safe:

• non-actionable FYI notes
• low impact with a clear workaround until morning
• change risk too high off-hours

With this, “not tonight” becomes a shared decision, not personal drama.

Micro scripts and a 7 day pilot

Scripts matter because they lower the social cost of being explicit when everyone is overloaded. They work best when they keep things respectful and give options. Also, it avoids the weird vibe where you sound like a policy memo.

• “Got it. No action tonight. Next update tomorrow 10:30.”
• “If this is urgent, please use the urgent lane, not multi-channel.”
• “Who is the owner for this, and what deadline are we aiming for?”
• “If I don’t hear back by 16:00, I’ll proceed with option B.”
• “Parking this for tomorrow. I’ll pick it up after standup.”

A small pilot beats a perfect policy. Pick 1 leak and plug it:

• response contract by channel plus a single urgent path
• end-of-day 3 line broadcast
• quiet-hours template as a default footer

Measure with 2 morning questions that match the standby feeling:

• did you wake up with an urge to scan status messages
• did sleep feel like 1 block, or like warm standby

Finally, a responsible guardrail. Work-design fixes can help. But if insomnia is persistent and impairing, for example ≥3 nights per week for ≥3 months, it’s worth talking to a clinician; CBT-I is often the first thing they suggest. Loud snoring with witnessed pauses in breathing or severe daytime sleepiness can also be a sign to check for sleep apnea.

Most teams don’t need a new habit. They need 1 boring rule that makes silence mean “safe.”