There are teams now in their fourth year of working together whose members have never been in the same room. They know each other's faces at a fixed resolution, in a fixed rectangle, in a fixed context: the meeting.
They are not strangers. But the bond is a working bond and nothing more, and working bonds have a specific failure mode. They hold perfectly while things go well and give way the moment something goes wrong.
What the corridor was actually doing
The informal moment gets discussed as if it were a perk: the coffee, the chat before the call, the walk to the lift. It was not a perk. It was infrastructure, and it was doing four jobs.
It carried information sideways. The thing you learned from someone in another team, unprompted, that stopped you building the wrong feature.
It let disagreements resolve cheaply. The two minutes after a meeting where you said "I wasn't sure about that" and it never became a formal position.
It made people into people. You found out someone has a sick parent, or plays in a band, or is terrified of the reorg. None of it relevant to the work. All of it changing how you read their next terse message.
And it built the reserve of goodwill you draw on when something goes badly. That reserve is not created during crises. It is created before them, in unstructured time.
Remote work removed all four, and almost nobody replaced them. A scheduled thirty-minute "virtual coffee" in a calendar invite is not the same mechanism. Structured informality is a contradiction, and people feel it.
Why more contact does not fix it
The instinctive response to weak cohesion is more meetings. It rarely works, and there is now evidence for why.
Brucks and Levav, publishing in Nature in 2022, ran a laboratory study of 302 participants plus a field experiment with around 1,500 engineers across five countries. Video calls significantly reduced creative idea generation compared with in-person meetings, and eye-tracking showed the mechanism: the screen narrows cognitive focus. You are looking at a face in a box, and your attention narrows to match.
Yang and colleagues, in Nature Human Behaviour the same year, studied 61,182 Microsoft employees through the shift to remote work. Collaboration networks became more static and more siloed, with fewer bridges between distant parts of the organization. People kept talking to the people they already talked to.
So the honest reading is this: adding calls increases contact and does not increase connection. They are different quantities.
What the body does that the screen cannot
Human trust is not primarily a cognitive assessment. It is built through synchrony: shared physical experience, matched rhythm, being in the same environment reacting to the same things at the same time.
Eating together does this. Being cold or hot together does this. Being slightly uncomfortable together, and getting through it, does it more than almost anything else. None of it survives translation to video, because video transmits face and voice and deletes everything else.
This is why the first physical meeting of a distributed team is so consistently reported as disproportionate. People describe it as finally understanding a colleague they had worked with for three years. Nothing new was said. The channel simply widened.
Getting more than a reunion
Most companies that recognise this book an offsite. A hotel, a conference room, a dinner, a team-building activity somewhere on the second afternoon.
It produces a good week and very little else, because it was designed as an event rather than as an intervention. Nobody established what was broken beforehand, so nothing was targeted, and nothing was measured afterwards.
Our approach starts the other way round. The diagnosis comes first, in two separate questionnaires, one completed by the company, one by the distributed team, so that the picture includes both what leadership believes and what people actually report. The gap between the two is usually where the cohesion problem lives.
The programme is then built on what came back. If the issue is that people have never been human to each other, the days are structured around shared physical experience rather than around sessions: ceremony, immersion in nature, a privately reserved cenote with no phone signal, meals cooked and eaten together. If the issue is that the informal channel died, we rebuild the habit deliberately and give the team the protocol to keep it alive at a distance.
And then we measure at ninety days, when the warmth has gone and only the structural change is left.
The alternative is a team that functions well and knows nothing about itself, holding together on process alone, until the first real pressure arrives.
LUNITA