Two years ago I sat in the back office of a business turning over something like one and a half million a year with fourteen people on the payroll. The owner had a videographer on £600 a month, in every Tuesday afternoon to film him. On the desk between us was a spreadsheet showing £41,000 of invoices out past ninety days, none of them chased since February, and a VAT bill due in eleven days that the account could not cover. He knew all of this. He could recite the debtor list from memory. What he wanted my opinion on was whether he was posting often enough.

I have thought about that afternoon a lot since, because the obvious reading of it is the wrong one. He was not lazy and he was not a fool. He had built the thing from a van. He was frightened, and ringing a man who owes you £11,000 makes the fear worse for the rest of the day, whereas filming yourself talking about your values makes it better for about four hours.

That is most of it, I think, and it has very little to do with marketing. Running your own business is lonely in a way that is hard to convey to anyone who has not done it. Posting is the only activity available at any hour of the day or night that produces immediate evidence you exist and are doing something. Chasing debt produces evidence too, of a sort nobody wants at nine in the evening.

There is a duller reason underneath it as well. You cannot see your reputation. You cannot see the customer who almost rang and then did not. Nearly everything that matters in a company reports late, in aggregate, and with a lag long enough that by the time you have the number you can no longer work out what caused it. Impressions and followers update by the hour and ask nothing of you except participation. Put a fast meaningless number next to a slow meaningful one and the nervous system attends to the fast one every time, which is what Charles Goodhart was getting at with monetary targets in the seventies. Once a measure becomes a target it stops working as a measure, because people start improving the number instead of the thing the number stood for. Hence a great many businesses with superb engagement and mediocre margins.

None of which amounts to saying social media does not work. It plainly does. It works for anyone whose product is bought on impulse, anyone whose personality is the product, anyone hiring, anyone selling into a market that finds things by scrolling. I have made money from it and so have people I respect. The argument is narrower than that. For most ordinary businesses the compulsion has come loose from the evidence, and it is the compulsion doing the steering.

The platforms understand this better than the people using them, and they built the payout to be erratic on purpose. Most of what you post does nothing. Then one lands, for reasons nobody can reconstruct afterwards, and there is a small flood of attention. Ferster and Skinner had the mechanism mapped by the late fifties. A pigeon fed every time it pecks gives up almost as soon as the food stops. A pigeon fed at random will carry on pecking thousands of times into a hopper that has given it nothing for hours. Fruit machines run on that arithmetic and so does your feed, and in the way this is usually described it is the business using the platform to condition an audience. The schedule is running on the man holding the phone. The reach he is pecking for is being throttled on top of that, since organic reach for business pages has been falling for more than a decade and now runs at a few percent of your own followers on most estimates. Two thousand followers, forty people. The free channel is a demonstration model for the paid one, which is a sensible thing for a platform to build and a strange thing to give your Sunday evening to.

Ten hours a week is a quarter of a working year. Nobody would knowingly assign a quarter of their year to an activity with no attributable revenue and yet the hours go somewhere. Ten hours a week is enough to ring every client who stopped coming since 2024 and ask why. It is enough to work out what happens at month four, which almost any service business could establish in an afternoon and almost none has. Putting your prices up six percent takes twenty minutes and the entire six percent lands on the bottom line, because nothing about delivering the work costs a penny more than it did the day before. There is no video in that.

Then there is the scrolling at eleven at night, which sets your own internal state, the cashflow gap and the difficult employee and the thing you got badly wrong in March, against other people’s advertising. Nobody posts their aged debtor report. And the cost is not really the hour spent, it is that you lie down afterwards and your mind will not stop running the odds on what you put up, so the business follows you into the bedroom in a form you can do absolutely nothing about at that time.

I am not above any of this. I have written posts at midnight that no client has ever mentioned and I have checked the numbers on them the next morning before I checked anything else. The only thing I have changed is the order of operations. Before I write anything now I ask what would be measurably different about the business in three years if I never posted again. Sometimes the answer is a great deal. Often it is nothing, and on a good week I go and do the boring profitable thing instead. On a bad week I write the post anyway.

He made the VAT bill, in the end. He borrowed it from his mother, which he mentioned to me about six months later and did not enjoy mentioning. He still films on Tuesdays and the videographer is on a retainer now. I am not sure he is wrong to. I am fairly sure he has never worked out what it is for.

Somewhere in the second week of August a certain kind of manager begins to sound uneasy. The pipeline has gone quiet. Three of the six people who can approve anything are unreachable. The client who was going to sign has an automated reply naming a date in September. The instinct that follows is nearly always the same. Treat the quiet as a loss of momentum, and counter it with a push, a nudge, a mildly passive aggressive message sent at half past nine on a Tuesday evening.

I think that instinct deserves closer inspection, because it is diagnostic. What an organisation does in August tells you far more about it than anything printed in its values document.

Take the available numbers first, and take them with some suspicion. The Global Payroll Alliance surveyed 1,132 workers a year ago this month and found 61 per cent admitting to a fall in productivity over the summer, roughly three in ten losing somewhere between half an hour and an hour of useful work a day, and another quarter losing more than an hour. It is worth being honest about what that survey actually asked, because it is not quite my subject. The reasons respondents gave were heat, tiredness, poor ventilation and an inability to hold their concentration. Nobody was asked about the absence of their colleagues. So it establishes that summer degrades output, which is useful, and says nothing whatsoever about what an emptying office does to a business. It is also self reported, and people are unreliable narrators of their own productivity in both directions. Hold it loosely.

What is not in dispute is the timing, and here Europe differs from North America in a way that matters. Europe does not stagger its leave. It closes.

The interesting thing is not the slowdown.

Most organisations shed something close to a third of their working capacity for four or five weeks and carry on functioning. Deadlines shift. Nothing catastrophic occurs. September revenue is rarely worse for it. If that is true, and in my experience it usually is, then the useful question is not how to defend momentum through August. It is what exactly the other eleven months are doing with all that surplus capacity.

My suspicion is that a good deal of it is consumed by internal weather. Meetings held to prepare for meetings. Approval chains with four links where two would serve. Work generated by the presence of people who need to be observed working. Remove a third of the staff and much of that becomes logistically impossible, and the organisation is pushed back onto whatever genuinely has to happen. August is not a slump so much as an audit nobody commissioned and nobody reads.

So far this sounds almost cheering. Here is where it stops.

Timetastic, a British leave management platform, went through the booking behaviour of something over 180,000 employees for its report on how the country actually uses its holiday. A quarter of workers reached the end of the leave year with more than three working weeks unused. Fewer than one in five took the whole allowance. And of the holidays booked through the platform during 2025, roughly a fifth were later cancelled, with the average request having gone in around four months ahead. People are not forgetting. They are booking the time, watching it approach, and then handing it back. The occupational psychologist working on that report kept arriving at the same explanation, which is guilt. Not guilt towards the employer, interestingly. Guilt towards colleagues.

That guilt is not irrational. It is arithmetic. Somebody absorbs the work. In a thin team an absence is not neutral, it is a transfer, and everyone can see who is receiving it. Which makes the summer quietly redistributive. The people most willing to take their entitlement are subsidised by the people least willing to take theirs, and the second group tends to contain your most conscientious staff. Run that for a few consecutive years and you have built a machine that systematically exhausts the people you can least afford to lose.

Then there is the question of what a holiday actually buys the person who takes it. The best evidence is a meta analysis by Jessica de Bloom and colleagues, published in 2009, which found a modest improvement in health and wellbeing over a holiday, an effect size of about 0.43, followed by a fade back towards baseline of almost identical magnitude once work resumed. A newer meta analysis updating the same literature to 2020, covering thirteen studies and roughly 1,400 people, is blunter still. Wellbeing improves during the break, and after the first week back the difference from where the person started is statistically indistinguishable from nothing.

A fortnight in Puglia buys you about seven days.

This gets routinely misread as evidence that holidays are pointless, which is not what it shows at all. It shows that a holiday is a suspension rather than a repair. Whatever produced the depletion is still there on the Monday morning, unaltered, and it reasserts itself at roughly the speed you would expect. The variable that matters is not the destination or the duration. It is psychological detachment, the term Sabine Sonnentag and Charlotte Fritz gave to the capacity to genuinely stop thinking about work, and of the four recovery experiences they identified it is the one most consistently associated with people actually getting better. Detachment is not the same as being physically elsewhere. A person can be in Corfu and entirely at their desk.

Sonnentag has also documented something she calls the recovery paradox, which is close to cruel. The higher the workload and the more sustained the pressure, the less able a person is to detach during time away. Recovery is hardest to obtain exactly where it is most needed. The employee who most requires August is the employee whose mind will not leave the building.

Which brings me to culture, a word I would happily see retired. Whatever culture is, it is not the thing described on the careers page, and it is not the summer party. It is the set of defaults that survive the absence of anyone enforcing them. August tests that directly. Take away most of the observers and see what remains. If work continues because people understand what matters and can decide without supervision, that is a culture. If everything stalls the moment the person who signs things goes to the Algarve, what you had was not a culture but a chain of command wearing one.

The retention consequence is the one I would ask leaders to take seriously, because it arrives late and looks unrelated to its cause. Resignations cluster in the late summer and early autumn, and recruitment activity in September and October is second only to the January surge. The standard explanation is that companies simply restart their hiring after the break, which is true and insufficient. The other half is that people do not resign at their desks. They resign on the third or fourth day of a holiday, somewhere around the point at which detachment finally arrives and they can see the job from the outside for the first time in a year. Nothing is announced. They go back, work the autumn, and hand in notice in October about a decision that was made in August with a paperback face down on a lounger.

So the summer risk to a business is not slower output in the weeks when half the office is away. It is the quiet reappraisal going on in all those absent heads, and the slow grinding of the people who stayed.

I have no elegant remedy, and I am wary of anyone who offers one. Two things seem worth saying. The first is that if you must intervene in August, intervene on cover rather than on output. Cover designed in advance, with explicit permission to let certain things simply wait, is the only mechanism I know of that stops absence becoming a tax on the conscientious. The second is that if your organisation genuinely cannot cope with a third of its people being away, the problem is not August. You have a concentration risk, and it will find you again in November under circumstances considerably less pleasant than a holiday.

There is a good deal I cannot tell you. The productivity surveys are self reported and commercially sponsored, which is not a reason to discard them but is a reason to hold them loosely. The recovery research rests on small samples, mostly white collar and mostly European. And I have seen nothing serious on what distributed work has done to any of it, which strikes me as a real gap, since the whole idea of a quiet office assumes an office.

What I will say is that the emptiness of the next few weeks is information, and most organisations throw it away. They spend August trying to make August look like March, and in doing so they miss the one month in the year when the business is not performing for anybody. Nobody is watching. The theatre has stopped. Whatever is still moving out there is the actual company.

What does your organisation look like when nobody is watching?

August gives you a rare opportunity to find out.

If you lead a team, use the next few weeks as a diagnostic. Look at what continues without intervention, what stops when one person is absent, and who quietly absorbs the work when others take their leave.

If you want an outside perspective, I work with organisations to assess the systems, behaviours and working conditions that sit underneath performance.

Talk to me about your organisation

I have been trying to work out when exactly people stop being able to think, and for most of them the answer is somewhere around half past two.

That is not a joke and it’s not really a metaphor. You can watch it happen. The morning version of somebody makes decisions they can still defend a week later. The afternoon version writes the email. Same person, same job, same set of problems, and a four hour gap in which something went.

Nobody in the wellbeing industry seems remotely interested in this. What they are interested in is Thursday’s resilience workshop. It runs for an hour, attendance is expected, the work it displaces is still due on Friday, and the invitation arrives with a photograph of a woman doing yoga on a jetty.

Read more

We talk about work and life as though they were two rooms with a door between them. The metaphor is convenient and almost entirely false. There is one body, one nervous system, one supply of attention, and none of it recognises the commute as a boundary. Whatever happens to you between nine and six is not left at the office. It walks through your front door wearing your face.

I want to be clear at the outset about what kind of argument this is, because I think it is usually smuggled in rather than stated. Everything below is empirical and I have tried to represent it honestly, including where it is weaker than I would like. But the conclusion I am driving at is a moral one. The state you are in when you come home is not a private matter. We are encouraged to treat dissatisfaction at work as a question of personal fulfilment, mildly self indulgent, something to resolve quietly on your own time. The research says it is nothing of the kind. It is transmitted. It is measurable in other people. And a number of those people did not choose your job.

That claim is only worth anything if the evidence carries it, so let me put the evidence first.

What the day does to the evening

Read more

Most people my age once carried a dozen phone numbers in their heads. Now they carry none, and the loss registers only in the rare moment the phone is dead and a number is needed. The same is true of directions. There are streets I have driven fifty times while following a small blue arrow, and I could not draw the route on a napkin to save my life. A faculty does not send a letter when it leaves. It thins, quietly, until the day you reach for it and find less there than you expected.

I want to make a case that this ordinary experience is the small, visible edge of something much larger, and that the same process is now working on the one capacity you can least afford to lose, which is your judgment. The argument is not that these tools make us stupid. That framing is lazy and mostly false. The argument is more specific and more uncomfortable. Every time you accept a correct answer from a machine without doing the work that would have produced it yourself, you take the output and skip the training. The output is useful today. The training was for the day the machine is wrong. And it will be wrong.

Consider the cleanest case we have, because it involves something almost everyone now does. A 2020 study at McGill University tested fifty regular drivers and found that people with more lifetime experience of satellite navigation had worse spatial memory when they were asked to find their way without it. That on its own proves little, because you might expect people with a poor sense of direction to lean on the device more. So the researchers went back three years later and retested part of the group. The people who had used navigation more heavily in the intervening years had declined more steeply. And the drivers who relied on it most had not started out with a weaker sense of direction. The tool was not compensating for a deficit. It was creating one. The lead researcher’s advice is the whole practical lesson of this essay in miniature. Use the device to reach somewhere new. Then switch it off and find your own way home.

Navigation is low stakes. Now raise them. The most mature version of this problem lives in the cockpit, where automation arrived decades before it reached the rest of us and where the consequences of skill loss are counted in lives. On the first of June 2009, Air France 447 fell out of the night sky into the Atlantic, killing all 228 people aboard. The proximate trigger was trivial. Ice blocked the sensors that measure airspeed, the autopilot did what it is designed to do in the face of unreliable data and handed control back to the pilots, and the pilots, who had spent their careers managing systems rather than flying by hand at altitude, pulled the aircraft into a stall and held it there all the way down. They were not incompetent; they were competent at the wrong thing. A senior training captain had warned years earlier that the industry was raising a generation he called the children of the magenta, pilots who could follow the coloured guidance lines on their screens beautifully and had quietly lost the reflexes underneath. Investigators keep circling the same paradox. The more reliable the automation, the less the human practises, and the less the human practises, the more catastrophic the rare moment when the automation steps aside. A United States Department of Transportation inspector general report put it plainly. The regulator had no adequate way even to assess whether pilots had kept their manual flying skills. The skill was assumed, and it was not there.

You might tell yourself that pilots and satellite navigation are special cases, one exotic and one trivial, and that thinking is different. But consider a study published last year in The Lancet Gastroenterology and Hepatology, a paper that ought to be far more famous than it is. Researchers looked at nineteen experienced endoscopists across four centres in Poland, each of whom had personally performed more than two thousand colonoscopies. These are not students. They are masters of a specific perceptual skill, spotting the small precancerous growths called adenomas on a screen, growths that are easy to miss and that kill people when they are missed. The centres introduced an AI tool that draws a box around suspected growths, and the researchers compared how well these doctors did on their own, without the tool, in the three months before it arrived and the three months after. In the before period the doctors found adenomas in 28.4 per cent of unaided examinations. In the after period the same doctors, working unaided again, found them in 22.4 per cent. That is a fall of six percentage points, a fifth of their detection rate, gone. This is the first evidence from ordinary clinical practice, rather than a laboratory, that exposure to AI can degrade the unaided performance of experts. A few months of having their attention guided was enough to erode the independent attention of people who had done this tens of thousands of times. And not one of them, I would wager, felt themselves getting worse. That is how this always works. The decline is silent.

If it can happen to a surgeon’s eye after three months, you should assume it can happen to your thinking. In early 2025 researchers at Microsoft and Carnegie Mellon surveyed 319 people who use these tools at work at least weekly and gathered 936 concrete examples of tasks they had done with AI. The finding that matters is not that people felt they were working less hard, though they reported exactly that across almost every kind of thinking, from comprehension to synthesis. The finding that matters is the relationship between confidence and effort. The more a person trusted the tool, the less critical thinking they applied to its output. The more they trusted their own ability, the more they applied. This describes a trap with a lock on the inside. The skill you need in order to catch the machine when it is wrong is judgment. Judgment is also the first thing to fade when you stop exercising it. So the people most exposed to a confident, fluent, plausible, and occasionally wrong answer are steadily losing the very capacity that would let them notice the occasions when it is wrong. Their work shifts from producing an answer to approving one, and approval, done without friction, is barely thinking at all.

There is even a signature of this in the brain, though I offer it with more caution because the work is newer. A group at MIT’s Media Lab had 54 people write essays while wearing electrode caps that record electrical activity. Some wrote with a chatbot, some with a search engine, some with nothing but their own heads. The people writing unaided showed the richest, most connected brain activity. The people using the chatbot showed the least. The search engine group showed between 34 and 48 per cent less neural connectivity than the unaided group. The chatbot group showed up to 55 per cent less. When the chatbot group were later asked to recall what they had just written, most could barely do it. After the first session, 83 per cent of chatbot users were unable to recall key details of their own essays. They remembered roughly a sixth of their own essays, against nearly half for the group that had worked unaided. And when the chatbot users were later moved to writing on their own, the disengagement lingered, as though the habit of not trying had outlasted the tool. This is a small study, not yet through peer review, and I would not build a cathedral on it. But it points in the same direction as everything above, and things that point in the same direction from cockpits, colons, offices and laboratories deserve a hearing.

Here is what I think is actually going on, stripped of the studies. Knowledge can be handed over. Judgment cannot. Judgment is not a store of facts you can look up. It is the residue left behind by having tried, having been wrong, and having felt the correction. It is built the way a callus is built, through repeated contact, and like a callus it softens the moment the contact stops. When you struggle toward an answer and get there, two things are produced. One is the answer. The other, invisible, is a small increment of judgment, a slightly better sense next time of what a good answer looks like, where the traps are, when something is quietly off. A machine can hand you the first for nothing. It cannot hand you the second at any price. You can only make that yourself, and only by doing the thing.

Which is why the machine being right is the problem, not the reassurance. If these tools were obviously and frequently wrong, we would stay sharp out of self defence, checking everything, trusting nothing. They are not. They are right often enough to earn a trust that quietly retires our vigilance. That is the trap. Competence in the tool breeds complacency in the user, and complacency is indistinguishable from progress right up until the day it is not. The bill does not come due on the ordinary tasks the machine handles well. It comes due on the rare, strange, high stakes case that sits outside everything the machine has seen, the case that most needs a human who has kept their judgment intact. That is precisely the case for which you will have traded your judgment away, one convenient correct answer at a time.

If you lead people, this stops being a private matter and becomes a question of what you are building. The evidence already laid out points to a troubling conclusion about organisations, without needing a separate study to confirm it. The endoscopy data shows that even masters of a craft lose their edge when a machine guides their attention. The aviation case shows that a training pipeline can quietly stop producing people who can perform the core skill when the automation fails. Extend that logic across any organisation that hands its junior work to machines. The difficult, fumbling early years in which competence is actually forged are precisely the years being automated away. You can raise this quarter’s output while quietly dismantling the mechanism that produced every capable person you currently rely on. Nobody will see it happen, because deskilling never announces itself. It shows up years later as a bench with no depth and a room full of people who can operate the tools but cannot tell when the tools are wrong.

None of this is an argument against the tools. It is an argument against using them thoughtlessly, which is the only way most people use anything. The correction is not abstinence. It is deliberate friction. Do the hard version first, then bring in the machine to check yourself, rather than starting from its answer and lightly editing what it gave you. Keep a set of tasks you always do unaided, the way a pilot keeps flying by hand even when the autopilot would do it better, precisely so the skill is there on the day it is needed. Treat verification as a real discipline that has to be practised and protected, not a box to tick at speed. And follow that researcher’s rule wherever you can. Let the device take you somewhere new, then switch it off and find your own way back. The struggle you are tempted to automate away is not waste. It is the training. It was always the training.

The reward and the cost arrive disguised as each other. Every time the machine is right and you did not have to think, it feels like a gift. It is actually a small withdrawal, made from an account you cannot easily refill, against a day you cannot see coming. The people who keep their judgment will be the ones who chose, again and again, to do the work they no longer strictly had to. Choose accordingly.

Share

Sources: The McGill navigation study is Dahmani and Bohbot, 2020. The Air France 447 investigation is BEA final report, 2012. The endoscopy study is Budzyń and Romańczyk et al., The Lancet Gastroenterology and Hepatology, 2025. The Microsoft and Carnegie Mellon survey is Lee et al., Proceedings of the 2025 CHI Conference. The MIT EEG study is Kosmyna et al., Media Lab, preprint 2025.

I used to believe that rest was simply what happened when work stopped. You finish the last email, close the laptop, and the recovery begins automatically, like a kettle cooling on its hob. It took me years to realise this is not true. It took reading the work of people like Sabine Sonnentag at the University of Mannheim to understand that my nervous system had no off switch. My body left the office, but my mind remained there, running simulations of conversations I would never have, replaying decisions I could not unmake. I was horizontal and yet entirely at my desk.

The research on recovery is now extensive enough to draw firm conclusions. Jessica de Bloom at Radboud University tracked what happens to people on holiday. Health and wellbeing rise during the trip, as you would expect. Then she measured what happened after. The benefits vanished within days of returning to work. In most of her studies, the effect had disappeared entirely within a week, and in almost none did it survive a month. Two weeks in Portugal produces a physiological dividend roughly as durable as a good weekend. This is not a failure of holidays. It is a failure of our model of recovery. We treat rest as a passive state when it is actually an active skill, and like any skill left untrained, most of us are terrible at it.

Read more

A friend of mine used to run operations for a fast growing company, and she once told me the exact moment she knew she was leaving. It wasn’t the day she handed in her notice. It was months earlier, in a meeting where her boss praised a colleague for rescuing a failing project, a project that same colleague had caused to fail in the first place. Nobody else in the room seemed to register the irony. She did. And something in her quietly closed.

That is usually how it goes. The exit rarely begins with a better offer landing in an inbox. It begins earlier, in some small moment nobody else notices, when a capable person sees the gap between what an organisation says about itself and what it rewards.

The standard explanation for why good people leave is almost always wrong, or at least incomplete. Pay, burnout, a rival poaching them, all of that happens, sure. But underneath it there is often something closer to what a philosopher might call an intolerable inconsistency. And the people least able to live with inconsistency happen to be exactly the people you most need to keep.

Read more

We have all encountered the person who protests too much about their own honesty. The insistence itself becomes grounds for suspicion. And yet, when we transpose this dynamic into corporate life, we somehow fail to recognise it for what it is. The organisation that blankets its offices with laminated declarations of integrity and collaboration is very often the one where integrity is a negotiable convenience and collaboration is a career liability. This is not merely hypocrisy, though hypocrisy is certainly part of it. It is something more structurally interesting. It is the predictable outcome of mistaking an advertising campaign for a cultural intervention.

The mistake runs deep. Most organisations treat culture as a communication problem. They believe that if they say the right words often enough, the behaviour will eventually follow. This gets things precisely backwards. Culture is not what you say. It is what you do when the pressure is on, when the cameras are off, when the right decision costs something tangible. It lives in the gap between the poster on the wall and the decision made at 4pm on a Friday. And the wider that gap, the more corrosive the culture becomes, because a stated value that is visibly not lived is not neutral. It is actively destructive. It teaches people that the real rules are different from the official ones and that understanding this difference is the primary survival skill.

Read more

There is a pattern hiding in plain sight across the working world, and once you see it you cannot unsee it.

The organisations with the most hostile internal cultures, the most punishing expectations, the most sophisticated machinery for silencing dissent, are not the ruthless profit-maximising corporations that feature in business school cautionary tales. They are the charities. The social enterprises. The purpose-led companies with the carefully worded values statements and the mission printed on the wall in a tasteful sans-serif font.

This is not a coincidence. It is a predictable psychological consequence of how human beings behave when they believe they are doing something important.

Understanding why this happens is not merely intellectually interesting. It is practically essential if you work in, lead, or are considering joining any organisation that leads with its values.

The Moral Licence Problem

Read more

A few years ago I sat in on a strategy meeting where the regional figures were presented as a tidy success story. Retention was up, complaints were down, and the only note of concern was a slightly soft number from one location, which everyone agreed was probably down to the weather, or a temporary staffing gap, or some other passing thing nobody felt the need to look into further. I happened to know two people who worked at that location. Neither of them was in the room. Both of them, if you had asked, would have told you exactly what was going wrong, because they had been telling their manager for months. Their manager had not passed it up, not out of malice, but because the last time someone in that role had raised something inconvenient, the response had been a polite request to focus on solutions rather than problems. So the manager found a solution. The solution was silence.

That is the story I keep returning to when I think about why organisations, full of intelligent and well meaning people, so often end up making decisions that look obviously wrong to almost everyone except the people making them. Nobody lied in that meeting. Nobody hid a report. The numbers on the slide were the numbers that had been collected, in good faith, by people doing their jobs. And yet the picture in the room bore almost no relationship to the picture on the ground, and the gap between the two had been built, patiently, one small accommodation at a time, by people protecting themselves from the cost of being the bearer of bad news.

I want to spend some time with this, because I think it is one of the most consequential and least understood failures in business, and because the cause is not technical. It is moral, in the sense that it concerns trust, and trust is not a soft virtue you bolt onto a strategy once the real work is done. It is the substrate the real work depends on.

Read more

There is a particular kind of management theatre that has become so commonplace in corporate life that most of us have stopped noticing how strange it is. A senior executive stands in front of several hundred people, or records a video message that will be watched at 1.5x speed, and delivers a set of financial imperatives with the gravity of someone announcing a military campaign. Volume growth. Positive mix. Margin expansion in hard currency. The audience nods. Someone in the third row writes it down. And then everyone goes back to their desks and does more or less exactly what they were doing before, because none of those words connect to any decision any of them will make that afternoon.

Fernando Fernandez, the chief executive of Unilever, has reportedly taken to greeting people not with hello but with a recitation of the company’s strategic priorities. Volume growth, positive mix, consistent growth margin expansion for profit growth in hard currency. Read that sentence again. Then ask yourself: if you were the person responsible for marketing Dove shower gel in, say, the Netherlands, what would you actually do differently on Monday morning as a result of hearing it? The honest answer is nothing. Because it tells you nothing. It is the linguistic equivalent of a motivational poster, except that motivational posters at least have mountains on them.

This is not a failure of communication in the ordinary sense. It is something more revealing than that. It is a window into a particular theory of human behaviour that senior leaders hold, often without realising it, and which the evidence suggests is almost entirely wrong.

Read more

Let us start with a fact that ought to stop you mid-scroll.

Seventy-one per cent of US chief executives report experiencing imposter syndrome in their current role. Not junior executives. Not people new to the job. Chief executives. The individuals sitting at the apex of some of the most complex organisations on earth, managing billions in capital and thousands of human lives, convinced that at some point this week someone is going to work out that they do not actually belong there.

Read more

The Monday morning ritual of telling yourself this week will be different, that you will finally catch up and move with focus, is not merely an innocent piece of self-motivation. It is a cognitive error with measurable consequences for your attention, your memory and the quality of your decisions. When you begin the week by pretending the slate is clean while ignoring the residue still clinging from the last, you guarantee that most of your present moments will be only partially lived.

The phenomenon has been studied in cognitive psychology as attention residue. In a series of experiments published in the journal Organizational Behavior and Human Decision Processes, researchers found that people need to stop thinking about one task in order to fully transition their attention and perform well on another. Yet it is difficult for people to transition their attention away from an unfinished task, and their subsequent task performance suffers as a result. The persistence of cognitive activity about a task even after you have stopped working on it and moved to something else is precisely what attention residue means. The research showed that not having finished a task before switching is a significant factor in how much residue you carry forward. But even completing a task is not always sufficient. Time pressure while finishing the prior task was found to be necessary for effective disengagement, because it forces the mind to let go rather than lingering in a state of unresolved vigilance.

Read more

There is a finding in organisational psychology that has never received the attention it deserves. A single person behaving badly inside a team, not catastrophically, not criminally, just reliably and corrosively badly, will reduce that team’s productive output by somewhere between thirty and forty percent. Let that number sit for a moment. We are not talking about a rogue executive committing fraud, or a manager who shouts. We are talking about the ordinary, unremarkable presence of someone who withholds effort, expresses cynicism, and treats basic norms of cooperation as optional. That person, embedded in an otherwise functional team, reliably destroys roughly one third of what that team could have produced.

The source is a 2015 meta-analysis from researchers at the University of New South Wales and the University of Washington, synthesising decades of studies on what the literature calls counterproductive work behaviour. The number is not contested. What is remarkable is how rarely it informs actual management practice.

The original laboratory evidence came a decade earlier. Will Felps and colleagues at the University of Washington placed deliberately negative participants, slackers, pessimists, and what the researchers neutrally termed jerks, into otherwise high functioning student teams. In every condition, performance collapsed. Creativity fell. Problem solving slowed. Team members began mirroring the negativity or withdrawing entirely. The paper’s title, How, When, and Why Bad Apples Spoil the Barrel, has since become a cliche, which is a shame, because the underlying finding is genuinely important and genuinely underused.

Read more

The constraints of corporate life frequently confine professionals to the home. Extended hours before screens erode the natural pathways through which human connection once arose. In this environment the question of companionship becomes central to overall functioning. Corporate wellness recognises that an employee is not an isolated unit of productivity. The state of mind and the quality of personal relationships shape focus, resilience and long term performance. Online dating enters this picture as a direct response to the very conditions many wellness programmes now seek to address.

Remote and demanding work patterns often diminish opportunities for organic social contact. Isolation creeps in. Loneliness correlates with heightened anxiety, lowered mood and diminished self esteem. These states in turn affect concentration, absenteeism and engagement at work. Progressive corporate wellness initiatives therefore attend to the full spectrum of human need including the social and relational dimensions. They acknowledge that sustained well being requires more than physical exercise or mindfulness apps. It demands attention to the human desire for intimacy and belonging.

Read more

Irving Janis came up with this idea in 1972. He had been looking at a string of policy disasters. The Bay of Pigs back in ‘61 was the big one. The people who planned that invasion were not fools. They had experience. They had degrees. They had access to real intelligence. So what went wrong? They handled dissent badly. Someone raised an objection. It got smoothed over. Someone expressed doubt. That got labelled as not being a team player. The plan kept moving on its own internal logic. Nobody stopped to check whether that logic still worked in the real world.

This same pattern turns up everywhere. Not just in governments. In your workplace. In your group chat. In your own head.

Think about the Challenger disaster. Engineers at Morton Thiokol saw a problem with the O rings in cold weather. They wrote it down. They talked about it. Then launch pressure built up. Schedules mattered. And they had seen O rings fail before without catastrophe. So the warnings got overridden. Not dismissed exactly. Just outweighed.

Read more

Most of us accept certain limits at work without really examining them. A quarterly target feels out of reach given current conditions. A product launch cannot happen in less than nine months because that is how things have always been done. Collaboration across time zones is dismissed as unworkable for creative teams. People state these things with the quiet confidence of someone reporting a fact. But look more closely. Most of these limits are not fixed by physics or economics. They are simply thresholds no one has crossed yet. That difference matters more than we usually think. It is the difference between small improvements and genuine breakthroughs.

Consider Roger Bannister. On 6 May 1954 at the Iffley Road Track in Oxford, a 25 year old medical student ran one mile in three minutes and 59.4 seconds. For years, experts including physiologists and coaches had insisted that the four minute mile lay beyond human capacity. The body could not withstand the strain, they said. The previous world record had stood at four minutes and 1.4 seconds since 1945. Progress had stalled. Bannister himself trained lightly while working hospital shifts. But he understood something the experts missed. The barrier was more psychological than physiological. Two fellow runners paced him through the early laps. He crossed the finish line and broke the record. Forty six days later, the Australian John Landy ran three minutes and 57.9 seconds. Within a few years, dozens of athletes followed. Today more than two thousand men have recorded sub four minute miles. The current world record is three minutes and 43.13 seconds, set in 1999. What changed was not human biology. What changed was the collective belief about what biology could do.

Read more

The constraints of corporate life frequently confine professionals to the home. Extended hours before screens erode the natural pathways through which human connection once arose. In this environment the question of companionship becomes central to overall functioning. Corporate wellness recognises that an employee is not an isolated unit of productivit…

Read more

Here is a fact about human beings that the self-help industry would prefer you never encounter. The people you admire for their discipline are almost never engaged in a heroic struggle against temptation. They have simply arranged their lives so that the right course of action requires less effort than the wrong one. The difference between the executive who reads before bed and the one who scrolls until his eyes burn is not a difference in moral fibre. It is a difference in the distance between the bedside table and the charging socket.

This sounds too simple to be true. That is precisely why it is worth attending to. We have built an entire culture around the worship of willpower. We tell ourselves that success belongs to those who can force themselves through another late night, who can resist every distraction, who can simply decide to focus and then do so without fail. This story flatters us. It suggests that our struggles arise from a personal failure of resolve rather than from any deeper mismatch between the way our minds actually function and the environments we have constructed.

The truth is less dramatic but more useful. Willpower is not a sturdy muscle that grows stronger with use. It is a fragile and inconsistent instrument, poorly suited to the sustained demands of modern work. The problem we face is genuine: we need greater clarity, depth and presence amid the constant pulls on our attention. The mistake is to treat that problem as a test of raw determination. You cannot out-will a poorly designed life.

Read more

The exhaustion arrives not as a dramatic collapse but as a fog that settles sometime in the middle of April. You carry on working. You keep turning up. The sharpness simply fades. Thinking slows. Discipline feels borrowed. A diffuse fatigue takes the place of the clear-headed drive that carried you through the first months of the year. This pattern returns with enough regularity that it deserves a name.

Call it the Q2 attention crash. It stems from no weakness of character. It arises as the predictable outcome of a fundamental mismatch between how human minds actually work and the demands we place on them.

The trouble begins with a mistaken picture of attention. Most people imagine it as a beam of light they can aim wherever they choose. That image leads them astray. Attention functions more like a finite resource, drawn down and renewed according to biological rules we are still mapping.

Read more