Whether consciously or not, we’ve all experienced the woes of team fragmentation.
The meeting that covers everything and decides nothing. The group chat that never stops. The talented people who are always busy and never quite available. And underneath it all, the slow erosion of the thing that makes a team worth having: the ability to think well together.
Fragmentation is not simply having too much to do. It is the cognitive cost of repeatedly disengaging, reorienting and reconstructing where you were. Every switch tugs and tangles the fragile thread that unites teams, and things begin to fray (Egner & Siqi-Liu, 2024).
In Article 2, we explored the ways we surrender attention as individuals. But individual choice can only take you so far. In teams, attention is shaped by collective norms: expectations, habits and unspoken rules that tell people what deserves attention and how quickly they should respond. Some protect the team’s capacity to think, while others tear it apart.
Norms are rarely a product of deliberate design. More often, they’re encoded as default settings through repetition and role modelling.
If you lead a team, you help write that code. You also have the power to rewrite it.
What follows is a map of six norms that commonly fragment team attention, grouped into three core team needs. For each, we offer a leadership “defrag”: a practical change to the patterns directing your team’s attention.
Team Rhythm
If you’ve ever been on a cruise, you’ll know that some onboard buffets never close – that smell of fresh pastry follows you down every deck, ever tempting just one more bite. Our access to information is just like that, except the information buffet living in our pockets is engineered to be far more irresistible than cruise ship food.
Fragmenting Norm: A narrow definition of good work
Most leaders carry an unspoken model of attention: Focused work is good work. More is always better. Anything else, such as a pause, walk, or long stare out the window, reads as attention leaking away. So we build accordingly: back-to-back calendars, instant-reply cultures, productivity measured by visible busy-ness.
As we discovered in Article 1, the brain does not run on focused attention alone. When external demands ease, the default mode network plays a greater role, supporting memory consolidation, imagining the future and reflective thinking (Menon, 2023).
A workplace of constant pings leaves little room for either state. People struggle to stay with demanding work, then fill the gaps that might have allowed ideas to settle.
Your best thinking happens when you finally give your mind some room.
The defrag: Design a rhythm, not a dial.
Treat unstructured thinking time as work rather than slack to be squeezed out. Protect empty space in the week. Defend it as a core deliverable.
Then, take an honest look at what you are modelling as a leader. If your own calendar is packed and your messages arrive at all hours, your team will read between the lines on what’s expected.
Fragmenting Norm: Conflating responsiveness with commitment
Meetings, messages, alerts and feeds create near-constant interruption. During core working hours, highly interrupted employees receive a meeting, email or chat notification about every two minutes (Microsoft, 2025). The average worker also receives around 117 emails and 153 Teams messages each weekday (Microsoft, 2025). Yet most people feel they can’t switch any of it off for fear of missing something or letting others down.
Message volume is only part of the problem. It’s the implication that fast replies signal commitment. No one wants to look like a poor team player, so everyone keeps every channel open all day, regardless of urgency or importance. Leaders reinforce this norm through what they notice, praise and respond to.
The defrag: Reframe commitment explicitly.
State the expected response time for non-urgent messages, measured in hours rather than minutes, and define one rarely used channel for genuine urgency. Reframe “do not disturb” as protecting productivity.
Then, live it. Delay messages that do not need to arrive after hours. Avoid praising instant responses as though speed were the same as value. Pay attention to the person who disappears for three hours and returns with the problem solved. Your behaviour tells the team which rules are real.
Team Safety
Teams struggle to think clearly while scanning for danger. Leaders shape that vigilance through the clarity they provide and the quality of relational repair they make possible.
Fragmenting Norm: Masquerading certainty, or ignoring its absence
Uncertainty about a possible threat can sustain vigilance and anxiety, even when the situation is outside anyone’s immediate control (Grupe & Nitschke, 2013). When stress becomes intense or prolonged, it can also impair the working memory and flexible thinking teams need to make good decisions (Arnsten, 2009). The result is a team scanning constantly for signals while becoming less able to think clearly about what to do next.
When facing uncertainty, the most damaging approach isn’t a leader who says nothing. It’s one who performs confidence they don’t feel. Usually, the intention is good: they want to steady the team and avoid adding to the anxiety. But people notice when the words and the atmosphere do not match. Instead of feeling reassured, they look for what has been left unsaid, intensifying threat-induced scanning loops.
The defrag: Offer clarity, not certainty.
You can rarely provide a definite prediction. You can always offer a clear logic: the direction we’re taking and why, the choices we’re making, and when we’ll know more. Name what you don’t know as readily as what you do.
People can work with uncertainty when they understand what is happening now and when they will hear more. What keeps them scanning is the suspicion that important information is being withheld.
Fragmenting Norm: Keeping the peace at any cost
For our ancestors, exclusion from the group could carry serious consequences, and the brain remains highly sensitive to social rejection. Experimental research has found that exclusion engages some of the same neural regions associated with physical pain (Eisenberger et al., 2003).
So when there’s unresolved tension, the mind keeps returning to it: the comment that felt off, the decision made without you, or the relationship that has suddenly gone cold. You may be sitting at your desk, but part of your attention is running in the background, watching for the next signal.
When a leader repeatedly chooses surface harmony over honest conversation, avoidance becomes another team default. The tension does not disappear. It moves into guarded behaviour, gossip and triangulation. People become preoccupied with where they stand, and have less capacity for the work in front of them.
The defrag: Make tension speakable and repair fast.
Normalise naming friction early, when it’s a conversation rather than a conflict, and model it by raising your own tensions first. Treat repair as urgent work, not awkwardness to outlast. Watch for signals: a person who has stopped contributing, an issue everyone talks around, conversations leaking into private channels.
Team Accountability
Teams think more clearly when people know what they are responsible for remembering, what should be easy to find and where human judgement remains essential. Without those agreements, people can end up carrying too much information in their heads while handing more important thinking to tools.
Fragmenting Norm: Expecting people to hold everything in their heads
Many teams carry an unspoken expectation that people should be across every detail. Everything must be recallable on demand. But human brains have a working memory of roughly four chunks at a time (Cowan, 2001).
When a leader regularly grills people for detail, the team adapts. People start memorising status updates they could easily look up because they do not want to be caught without an answer. Limited working memory gets spent on appearing across it all, rather than thinking deeply about what matters.
Offloading memory storage is smart use of resources. We naturally remember where information can be found rather than the information itself (Sparrow et al., 2011), a feature of how cognition is distributed across people and tools (Risko & Gilbert, 2016). Deep knowledge is still valuable, but the balance should be agreed rather than left to anxiety.
The defrag: Name what information is critical to immediate recall.
Agree what must be known deeply, what only needs to be findable, and who owns what. Then watch what your own questions reward: grilling for detail teaches the team that memorised breadth matters more than depth of thought.
Fragmenting Norm: Ceding judgement to tools
Offloading storage allows the mind to think, but under pressure, teams may offload the thinking too, handing problem framing and judgement to AI to save time. In a 2025 survey of 319 knowledge workers, greater confidence in generative AI was associated with less critical engagement, while greater confidence in one’s own ability was associated with more (Lee et al., 2025).
A polished, fast answer does not guarantee a well-considered solution, particularly if no one questions how the problem was framed, what may be missing or whether the answer fits the context.
The defrag: Be explicit about which work must stay human.
Normalise “think first, prompt second”, using AI to stress-test thinking rather than replace it.
Your guiding principle is to offload information storage and retain responsibility for judgement. By getting facts and status updates out of your team’s heads and into easily retrievable systems, you free cognitive capacity for the human work that still requires ownership: framing, synthesis, judgement and systems thinking.
When deep thought is required, resist the pull toward speed. Give people permission to sit in the discomfort of not knowing a little longer.
Ask people to explain how they reached a conclusion, not only to present the output. The questions you ask become part of the team’s code: they teach people what kind of thinking will be expected next time.
Most fragmenting norms were never deliberately chosen. They accumulated through pressure, repetition and whatever seemed easiest at the time. Eventually, the workaround became the way the team worked.
Defragging starts by examining those defaults and rewriting the ones that no longer serve the team. Where does the working rhythm keep getting interrupted? What leaves people scanning for information or relational threat? What are people expected to carry, and what have they stopped thinking through for themselves? Then, dismantling what no longer serves, and practising the conditions you want to keep.
When everyone has access to the same tools, the same information, the same external pressures, the rare advantage is a team that can still think well together.
This is the third of a three part series on Defragmentation:
Attention by Design
01. SYSTEM A systemic lens on how we fragment and the stakes.
02. SELF Where you’re surrendering attention and how to start taking it back.
03. TEAM The conditions that fragment team attention and how to dismantle them.
References
Egner, T., & Siqi-Liu, A. (2024). Insights into control over cognitive flexibility from studies of task-switching. Current Opinion in Behavioral Sciences, 55, 101342. https://www.sciencedirect.com/science/article/abs/pii/S2352154623000967?via%3Dihub
Arnsten, A. F. T. (2009). Stress signalling pathways that impair prefrontal cortex structure and function. Nature Reviews Neuroscience, 10, 410–422.
Cowan, N. (2001). The magical number 4 in short-term memory. Behavioral and Brain Sciences, 24(1), 87–114.
Eisenberger, N. I., Lieberman, M. D., & Williams, K. D. (2003). Does rejection hurt? Science, 302(5643), 290–292.
Grupe, D. W., & Nitschke, J. B. (2013). Uncertainty and anticipation in anxiety. Nature Reviews Neuroscience, 14, 488–501.
Lee, H.-P. et al. (2025). The impact of generative AI on critical thinking. CHI Conference on Human Factors in Computing Systems.
Menon, V. (2023). 20 years of the default mode network. Neuron, 111(16), 2469–2487.
Microsoft. (2025, June 17). Breaking Down the Infinite Workday. Microsoft WorkLab. https://www.microsoft.com/en-us/worklab/work-trend-index/breaking-down-infinite-workday
Risko, E. F., & Gilbert, S. J. (2016). Cognitive offloading. Trends in Cognitive Sciences, 20(9), 676–688.
Sparrow, B., Liu, J., & Wegner, D. M. (2011). Google effects on memory. Science, 333(6043), 776–778.
