Collaboration Theater
Why the teamwork you built is destroying the thinking it depends on
Last year, in “From Leadership to Teamship,” I argued that companies had outgrown the lone-genius CEO. Move from control to collaboration, I said. Continuous engagement over periodic check-ins.
I still believe the diagnosis. But I left out the dosage. And in medicine, a correct diagnosis with the wrong dosage is called poisoning.
Most organizations picked the cheapest possible interpretation. More channels. More meetings. More instant replies. More people copied “for context.” Now everyone is included, informed, consulted, and cognitively shredded. I asked for teamship. What a lot of companies built instead was collaboration theater: optimizing for the visible signals of coordination while quietly destroying the cognitive conditions that make coordination worth anything.
Here is the argument I should have made the first time, and it is not “protect deep work” or “have fewer meetings.” It is that most companies have fused two things that should never have been fused: information transparency and attention access. They made everyone informed by making everyone interruptible, then confused the interruptibility with the teamwork.
That fusion is the root. Everything else is symptom.
The Fusion Problem
Want everyone aligned? Put them in the channel. Want transparency? Make everyone available. Want context? Copy them on the thread.
Every one of these moves conflates knowing with being reachable. But information can be transparent without being pushed. Decisions can be visible without requiring attendance. Context can be shared without seizing attention. The difference between a well-organized shared dashboard and a 47-message Slack thread is not the information content. It is the delivery mechanism. One waits for you. The other hunts you.
Cal Newport calls the underlying workflow the hyperactive hive mind in A World Without Email: work organized around ongoing, unstructured, real-time messaging. It feels efficient because every question can be asked immediately. It feels collaborative because the conversation never stops. And it generates catastrophically low returns on the one resource knowledge work actually depends on, the sustained cognitive capacity of the brains you employ.
Gloria Mark’s research in Attention Span measures the cost. By 2021, knowledge workers switched attention every 47 seconds. Once interrupted, it takes an average of 25 minutes to return to the original task. And 44 percent of those switches are self-initiated. No ping. No notification. People interrupt themselves, because the always-on environment has trained their brains to expect constant input. The hive mind gets inside your head.
This is not a discipline problem. Adam Gazzaley and Larry Rosen explain in The Distracted Mind that our brains evolved powerful executive functions, the ability to set goals and plan, but far weaker cognitive control, the ability to stay on task despite interference. Our primate ancestors needed to be distractible. The rustle that pulled attention from foraging was occasionally a predator, and the ones who ignored it became lunch. Slack notifications exploit the same neural pathway. The brain cannot tell the difference.
So when you tell someone to “just focus harder” while running them inside an always-on communication environment, you are asking them to override neural wiring that kept their ancestors alive. The instruction is technically possible and practically useless.
The problem is not that your people lack discipline. It is that you are taxing attention through the same channel you use to share information, and the tax is invisible while the information flow is extremely visible.
The Case Against My Own Argument
Now here is where this gets genuinely difficult, because “protect deep work” has a flaw that most productivity writers either ignore or wave away. And the flaw nearly lost a war.
Stanley McChrystal ran into it in Iraq. In Team of Teams, he describes how the most elite military force on earth was losing to an enemy with fewer resources, worse training, and no org chart. The US military had optimized for efficiency. Each unit operated with precision within its own domain. Information flowed up, decisions flowed down, and every team protected its operational focus.
Al-Qaeda in Iraq operated on shared consciousness. Every node saw the full picture. Information moved laterally. Adaptation was instant because nobody was waiting for a briefing cycle.
McChrystal’s conclusion should make anyone arguing for “protected thinking time” deeply uncomfortable: the problem was not too much communication. It was that information had been siloed by the very structures designed to protect operational focus. By giving each team the space to work without distraction, the military had built an organization that could not learn fast enough to survive.
He dismantled the silos. He built radical transparency. Everyone saw everything, all the time. And it worked.
If McChrystal is right, then the “collaboration theater” argument has a hole in it large enough to drive a tank through. Maybe all those Slack channels and cross-functional syncs are not theater at all. Maybe they are the messy, imperfect, annoying mechanism by which a complex organization maintains the shared consciousness it needs to adapt. Maybe the person complaining about too many meetings is the organizational equivalent of McChrystal’s siloed units: locally focused, globally blind.
That objection is strong enough that I cannot dismiss it. And I spent several drafts trying to before admitting it was winning.
The Paradox That Resolves It
Here is what I missed in those earlier attempts.
McChrystal’s shared consciousness did not mean that every soldier was interruptible by every other soldier at all times. It meant that every soldier could see the full operational picture when they chose to look. The daily Operations and Intelligence briefing was not a 47-message chat thread. It was a structured, time-bounded, radically transparent information event. Between those events, operators operated.
In other words, McChrystal solved the information transparency problem without creating an attention access problem. He increased shared consciousness while protecting cognitive capacity. He did both.
Most companies cannot imagine this because they have welded information sharing to communication frequency so completely that the two feel like the same thing. When a leader says “I want more transparency,” the organization hears “make everyone reachable.” When a team says “we need alignment,” someone schedules a meeting. The tool for sharing information is the same tool that seizes attention, and nobody has thought to separate them.
This is the real argument, and it is not “protect focus” versus “share everything.” It is that the organizations that win will be the ones that solve both problems simultaneously. Radical information transparency. Fierce attention protection. Not a compromise between them. Both, by design.
In 1981, the FAA formalized the sterile cockpit rule after a series of crashes caused by distracted flight crews. Below 10,000 feet, only flight-essential conversation is permitted. Above cruising altitude, talk about whatever you want. The FAA did not reduce information availability. Crew scheduling, weather updates, and operational data were all still accessible. They just could not be delivered in a way that interrupted the cognitive work of landing the aircraft.
Knowledge work needs the same principle, adapted for the messier reality that you cannot always predict which phase you are in. The fix is not rigid scheduling. It is defaults. If something can wait four hours without material harm, it does not enter the real-time channel. If a decision needs more than three people, it starts with a written brief. If someone is in a declared work block, the interruption threshold rises to genuine dependency, customer harm, or regulatory deadline. Not “while I have you.” Not “quick question.” Not “just looping you in.”
These defaults sound obvious. They are also absent from almost every organization I have worked with. Designing communication protocol feels less urgent than responding to the next ping. Which is itself a symptom of the disease.
What This Changes About Teamship
The missing sentence from my earlier article is short: continuous engagement does not mean continuous interruption.
The Slack Threshold I wrote about in “Optimization is the New Fragility” applies here. There is a point where removing communication friction genuinely improves coordination. Past that point, you are cannibalizing the cognitive slack that makes coordination valuable. And crossing that threshold feels like progress, because the visible metrics keep improving right up until nobody has had an original thought in weeks.
The organization that rewards responsiveness over reflection will, over time, select for people who manage interruptions gracefully rather than people who solve hard problems well. Those are occasionally the same people. They are not always the same people. And the ones you lose are the ones whose thinking made the collaboration worth having in the first place.
The 15-Minute Attention Tax Audit
The Mode Mismatch (5 min)
List the five most important outputs your team produced last week. Not activities. Outputs. For each, mark the cognitive mode it actually needed: real-time coordination, structured rhythm, or shielded deep work. Then mark the mode it got. If three or more were handled in a noisier mode than they required, your defaults are degrading the work.
The Theater Ratio (5 min)
Pick one decision your team made via chat last week. Count total messages and total people involved. Now count how many messages contained new information that actually changed the decision. If fewer than a third moved the needle, the rest was performance. Could the same decision have been made with a written brief and one 20-minute conversation? If yes, you are paying a coordination tax for the privilege of looking busy together.
The Sterile Cockpit Test (5 min)
Pick one two-hour block this week and make it genuinely sterile. No meetings. No chat. No “quick question.” Leave one emergency path open and define what qualifies before the block starts, so nobody gets creative with the definition. At the end, count attempted interruptions versus genuinely urgent ones. If urgent is near zero and output rose, expand the block. If something genuinely breaks, good. You found where real-time coordination actually belongs.
Kill rule: If your team cannot name a single business outcome that would worsen from a four-hour delay on non-incident messages, kill the expectation of immediate replies.
Questions Worth Sitting With
When was the last time you interrupted someone’s deep work for something that, if you are honest, could have waited until tomorrow? What did that interruption cost them that you will never see, because the cost is invisible and your comfort was immediate?
If you audited every message your team sent last week, what percentage required a response within the hour, and what percentage just felt urgent because the notification was there, blinking, exploiting neural wiring that cannot distinguish a Slack ping from a predator in the grass?
Who on your team has quietly stopped doing their best thinking because your culture has taught them that being reachable is more valued than being right? Do you even know who they are? And what does it mean that your system has no way of finding out?
What would it actually take to build an organization where information is radically transparent and attention is fiercely protected? Not as a slogan. As a system. What would you have to give up?