How long should a team rest after crunch or a big launch?

There's no universal number, just as there's no single safety-stop formula. What matters is the *slope* of the return, not a fixed duration. A practical floor is two to three deliberately low-load days immediately after you ship — cleanup, documentation, and slack instead of a hard restart — followed by a gradual ramp back to normal velocity. The key is that the decompression happens right after surfacing, not weeks later.

Why do I feel exhausted after a deadline ends instead of during it?

For the same reason the bends appears after a diver surfaces, not at depth. During crunch you run on adrenaline and a shared goal, which mask the strain. When the pressure comes off, the accumulated load finally registers — often a week or more later. DAN notes decompression symptoms can appear "up to 12 hours after surfacing"; team-level post-launch burnout lags even further, which is why it's so often misattributed.

Isn't the safety stop just a vacation?

No — and the difference is the whole point. A safety stop happens *on the ascent*, immediately, while the pressure is still coming off. A vacation a month later is a hyperbaric chamber: treatment after the bubbles have already formed. Both have value, but only the immediate, deliberate pause prevents the injury. For sustainable pace, timing beats generosity.

How do remote teams run a safety stop without an office?

Co-located teams decompress naturally — the launch ends and the room exhales. Distributed teams have to make that surfacing deliberate. Run one shared, synchronous debrief where the team marks the end of the push together, ideally on video around a live canvas so everyone ascends at once instead of drifting back to Slack alone. Then protect a light-load window before the next cycle.

What's the difference between the bends and a meeting hangover?

A meeting hangover is daily cognitive fatigue from back-to-back calls — a depth-and-frequency problem you solve with buffers between meetings. The bends at work operates at the level of the cycle: the delayed injury from surfacing too fast after an intense push. One is about your Tuesday; the other is about how you come up from your quarter.

The Safety Stop: Why Teams Get the Bends After Crunch

The condition has a beautiful name and an ugly history. Divers call it the bends. Engineers called it caisson disease, because of how it was discovered. As Wikipedia documents, the term was "introduced in the 19th century, when caissons under pressure were used to keep water from flooding large engineering excavations" — the pressurized chambers used to sink bridge supports. During the construction of the Eads Bridge, "15 workers died from what was then a mysterious illness." On the Brooklyn Bridge, it "incapacitated the project leader Washington Roebling." The men crawled out of the pressure bent double in pain. Hence the name.

Here is the part that should stop you cold: none of them were hurt by going down. They were hurt by coming up too fast. And that is exactly why the safety stop — the diver's discipline of pausing on the way to the surface — is the most useful idea your team has never applied to itself.

Because your team has its own version of the bends. It just shows up a week after the launch, when nobody is looking for it.

The injury isn't the depth — it's the ascent rate

Spend time under pressure and your body absorbs nitrogen. The deeper you go and the longer you stay, the more you take on. That part is intuitive, and it is not the dangerous part. The danger is in how you leave.

Divers Alert Network, the leading dive-safety organization, puts the mechanism plainly: "During a dive, the body tissues absorb nitrogen... in proportion to the surrounding pressure," and "if the pressure is reduced too quickly, the nitrogen may come out of solution and form bubbles in the tissues and bloodstream." Wikipedia is even more direct about the controlling variable: "The faster the ascent from depth or ascent to altitude, the greater the risk of developing DCS."

Read that again, because it is the whole argument. The risk is not a function of how deep you went. It is a function of *how fast you came up*. A diver can spend an hour at depth and surface perfectly fine — if they ascend slowly enough for the gas to leave their tissues gradually. The same diver can do an easier dive and get badly hurt by bolting to the surface.

Now translate it. The "depth" is your crunch — the launch week, the incident response, the quarter-end push, the hardest stretch of a hard project. Teams obsess over that part. They ask whether the sprint was too long or the deadline too aggressive. Those are depth questions. They are not the ones that injure people.

The injury comes from the ascent. From going 0 to 100 and then back to 0 with no transition — from a team that ships at midnight on Friday and is expected to be at full velocity again Monday at nine. The pressure didn't hurt them. The speed of returning to normal did.

Why the bends shows up after you surface

The cruelest feature of decompression sickness is its timing. You feel fine at depth. You feel fine when you climb out of the water. Then it arrives.

DAN notes that "signs and symptoms usually appear within 15 minutes or up to 12 hours after surfacing." The statistics on Wikipedia are sharper still: in "more than half of all cases, symptoms do not begin to appear for at least an hour," and the onset curve runs 42% within one hour, 60% within three, 83% within eight, and 98% within 24 hours. The damage is booked during the fast ascent. The bill is delivered later — and because it is delayed, the diver rarely connects the pain to the mistake that caused it.

This is precisely what happens to teams, and it is why post-launch burnout is so badly misdiagnosed. The exhaustion, the resignation letter, the suddenly-cynical engineer — these almost never appear during crunch. People are running on adrenaline and a shared deadline; they feel *great*, the way a diver feels great at depth. The symptoms surface two or three weeks later, decoupled from their cause. By then leadership has moved on. So the attrition gets blamed on "culture," on "the market," or on the person who quit — when the real culprit was an ascent with no stop in it.

If you want the most extreme proof that workplace harm is delayed and cumulative, look at the hours research. The World Health Organization and International Labour Organization found that long working hours contributed to "745 000 deaths from stroke and ischemic heart disease in 2016." Working 55-plus hours a week is associated with a "35% higher risk of a stroke" and a "17% higher risk of dying from ischemic heart disease." Nobody dies of a busy Tuesday. They die years downstream, when the bubbles that formed under pressure finally do their work.

Burnout follows the same delayed arc. The WHO now classifies it as an occupational phenomenon — "a syndrome conceptualized as resulting from chronic workplace stress that has not been successfully managed." In plain terms: it builds up from stress the workplace never let people off-gas. Note the word *chronic*. It is the accumulation, not any single dive, that breaks people.

What a safety stop actually is

So divers built a fix into the ascent itself. Near the end of almost every recreational dive, you pause. PADI, the largest dive-training agency in the world, describes the standard: "stop at 5 metres / 15 feet for three minutes." The purpose, in their words, is "to give your body time to release that nitrogen slowly" — "as opposed to forcing the nitrogen out too quickly which can create bubbles and lead to decompression sickness."

Three minutes. That's the whole intervention. Not a shorter dive, not less depth — a deliberate, scheduled pause on the way up that costs almost nothing and prevents almost everything.

Two things about the safety stop are worth stealing exactly.

The first is intellectual honesty about the number. The precise "three minutes at fifteen feet" is a conservative convention, not a law of physics — different agencies cite slightly different depths and durations. The exact figure was never the point. The point is the *shape of the curve*: gas has to leave on a gradient, and you cannot rush a gradient. Your team's stop doesn't need a perfect formula either. It needs to exist and to be slow.

The second is the timing, and this is where most "wellness" advice gets it backwards. A safety stop is not a vacation you take a month later. By the time the bubbles have formed, the stop is useless — that's a hyperbaric chamber, not prevention. The decompression has to happen *on the ascent*, immediately, while the pressure is still coming off. The team equivalent of a safety stop is not "we'll do an offsite next quarter." It is the deliberate, protected, low-load window in the days right after you surface. Miss that window and a later break, however generous, can't undo the damage.

This is a different claim from the one about recovering between back-to-back meetings, which is about daily cognitive load. The safety stop operates at the level of the *cycle* — the push, the launch, the incident — and its logic is rate, not rest.

How to run your team's safety stop

You don't need a program. You need a built-in pause that obeys the ascent-rate rule. A few moves that work:

Schedule the stop before the push, not after. Divers don't decide whether to do a safety stop when they reach fifteen feet; it's part of the dive plan. Put the decompression window on the calendar the day you commit to the crunch, so it can't get optimized away in the relief of shipping.

Surface together, not alone. This is the part remote teams get most wrong. A co-located team has natural decompression — the launch ends, people exhale in the same room, somebody says "go home early." Distributed teams have none of that. The push just dissolves into the next thread, and everyone surfaces in isolation. So make the stop a shared, synchronous moment: a single live debrief where the team off-gases together — names what happened, captures what to keep, and explicitly marks the end. A working session on Coommit, with everyone on video around a shared canvas, turns that surfacing into one collective ascent instead of a dozen lonely ones. The artifact matters less than the shared exhale.

Ramp the load back gradually. The cardinal sin is the instant return to full velocity. After a major push, the next few days should be deliberately light — cleanup, documentation, slack in the schedule — before normal pressure resumes. You are managing the *slope*, not skipping the work.

Watch for symptoms the following week, not the launch week. Crunch week looks fine by design. Put your attention two and three weeks downstream, when the bends actually shows up, and treat a spike in cynicism or a surprise resignation as a decompression failure, not a personnel problem. This is the opposite instinct to the golden hour of fast response — some problems demand immediate action, but recovery demands the patient ascent.

The default working environment offers no stop at all. Microsoft's 2025 Work Trend Index found that "80% of the global workforce... say they're lacking enough time or energy to do their work," that employees are "interrupted every 2 minutes" for "275 interruptions a day," and that nearly half "(48%)" — and a majority of leaders — "say their work feels chaotic and fragmented." It's the steady-state version of the cost of context switching: a workforce permanently at depth, never ascending, never stopping. The bends, at that point, isn't a risk. It's the operating model.

The stop is non-negotiable, not a reward

A diver does not earn the right to skip the safety stop by being experienced or by being in a hurry. The stop is part of the dive. Skipping it doesn't make you tough; it makes you the next case study.

Teams treat decompression as a luxury — something you get to do if the quarter was good and the calendar allows. That has the logic exactly inverted. The stop is the cheapest insurance you will ever buy against your most expensive problem: Gallup estimates the cost of replacing one person at "one-half to two times the employee's annual salary," and pegs voluntary turnover as a trillion-dollar annual loss for U.S. business. Three protected days against that is not generosity. It's arithmetic.

You can't keep your team from going deep — depth is where the work is. But you control the ascent. Build the safety stop in, make everyone surface together, and slow the climb. Your team won't get the bends from the crunch. They'll get it from how you bring them up.