Thursday, October 15, 2026
Microsoft required three days a week for Puget Sound employees starting in February. UPS mandated five days for its thirty-six hundred corporate workers. The mandates are real. The compliance question is more complicated.
Meetings starting after 8pm rose 16 percent in a year, sixty percent of meetings are called on the spot, and companies that deleted their calendars entirely got worse rather than better. The packed calendar and the empty one describe the same unresolved question, which is who is allowed to end an argument.
On the Miami cockpit recording a pilot says the aircraft is going too fast, and says it again, about a hundred seconds before the crash. Aviation spent forty years building procedures for exactly this gap, because it turns out that identifying a problem and having it acted on are different events with different causes.
Aviation has a name for the moment when continuing and abandoning are no longer both available, and a rule that forces the choice before it arrives. Most organisations have neither. They discover the boundary by crossing it, usually while trying to be prudent.
Jaguar Land Rover has opened a voluntary programme covering about 4,000 salaried and management roles. A voluntary round is humane, cheaper to run and legally cleaner than the alternative. It is also the one method of reducing headcount in which the employer does not decide who goes.
The advice to delegate more is sound and incomplete. A manager who has never done the task cannot specify it, cannot judge the output, and cannot tell whether the person is struggling.
Information travelling up an organisation is edited at every layer by someone with an interest in how it reads. Skipping a level is the cheapest correction available, and most executives run it badly.
Policies accumulate because each was justified at the moment it was written. Almost none are ever removed, and the resulting rulebook governs a company that no longer exists.
Companies collect complaints diligently and route them to resolution rather than to comprehension. The signal that would have prevented the next hundred is discarded at the moment it is answered.
Every company has dashboards nobody opens, and every company responds by redesigning them. The reason they go unread is that no choice is waiting on what they say.
Companies retired the annual appraisal for good reasons and assumed ongoing conversation would fill the gap. Where nobody built the replacement, employees now go a year with no record of how they are doing.
The reward for being reliable is more of everything, and most of it is work that did not need this person. Companies measure headcount and never measure who is doing what.
Companies spend heavily to attract people and hand them a laptop and a login. The departures that follow eighteen months later were mostly determined in the first three weeks.
Most strategy decks describe a future the company would like and a plan for arriving there. A strategy names what the company will not do, and almost none of them do.
Co-leadership is being used again to bridge a succession, retain a rival or cover a gap nobody wants to name. It works where the split is genuine and fails where it was a way of avoiding a choice.
Most reorganisations are announced because a specific, uncomfortable decision was available and nobody wanted to make it. The chart changes, the decision waits, and everyone is busy for two quarters.
Boards have become disciplined about naming a successor for the chief executive. Two levels down, where most operating knowledge actually sits, the same companies have no plan and frequently no list.
Every firm removing junior work is making a decision about its capability in 2031, and almost none of them are making it on purpose. The right question is not what the work costs but what it teaches.
Organisations have become very good at running trials and no better at ending them. A pilot that produces a clear positive result is the one most likely to be extended indefinitely, because extension is the decision nobody has to defend.
The flattening that removed a layer of management left survivors with teams of fifteen and twenty. Organisations are now setting explicit spans, and discovering the number they can defend is smaller than the one they have.
A quarterly ritual built around a hundred-page deck produces informed directors and almost no useful challenge. The format is the problem, and it is entirely within the board's power to change.
Companies measure everything except the forty percent of payroll spent talking about work, and the audit they refuse to run would explain more than any engagement survey.
After several years of celebrated flattening, organizations are rediscovering what middle management was for, and rebuilding it under new names.
An improvised job created for trusted individuals is acquiring job descriptions, levels and career paths. What it is becoming depends heavily on what problem it was created to solve.
Organizations keep buying infrastructure to answer questions they have never agreed on the wording of. The disagreement is not technical, which is why the technical spending does not resolve it.
Disclosure laws were written to affect job postings. Their most significant effect has been inside the building, on conversations managers were never trained to have.
Speed has become the default virtue in hiring. It is the wrong one, and the companies quietly ignoring it are building better institutions.