A strange thing is happening to management.

Companies keep talking about how important managers are while simultaneously eliminating them, widening their spans of control, giving them more responsibilities, asking them to lead through more uncertainty, and wondering why so many of them seem exhausted.

This sounds contradictory until you realize we may be using one word to describe a job that no longer exists in one coherent form.

The manager is still here. The management bargain may not be.

For a long time, becoming a manager meant accepting a fairly understandable trade. You took on more responsibility, but you also received more authority, status, compensation, organizational access, and control over how work got done.

Those things traveled together often enough that “manager” meant something.

Increasingly, they don’t.

CLARITY

What becoming a manager used to mean

The traditional management job bundled together several things.

You supervised people. You allocated work. You evaluated performance. You had some control over resources. You translated direction from above and made decisions for the people below you.

And importantly, increased accountability generally came with some increase in authority.

Not perfectly, obviously. Organizations have been finding creative ways to make managers responsible for things they cannot control since approximately the invention of the meeting.

But the bargain was legible.

Move up the hierarchy and you would probably manage fewer tasks yourself while gaining more responsibility for the work of others. Keep moving up and your role would become increasingly strategic.

That progression helped make the organizational ladder make sense.

Now look at what is happening.

Gallup found that the average number of direct reports per manager in the U.S. increased from 10.9 in 2024 to 12.1 in 2025, nearly 50 percent higher than when Gallup first measured it in 2013. Gallup attributes part of the increase to companies consolidating middle-management positions and placing more employees under the managers who remain.

Deloitte reports that U.S. employers were advertising 42 percent fewer middle-management positions at the end of 2024 than they were in spring 2022. At the same time, organizations still need somebody to coach employees, resolve problems, translate strategy, make judgment calls, coordinate work, manage change, and increasingly figure out what humans should do when AI can do part of the job.

You can eliminate the manager.

You cannot eliminate the work of management.

When organizations remove a layer, those responsibilities have to go somewhere. Some move upward. Some move downward. Some get automated. And a surprising amount simply gets added to the job description of whoever survived the restructuring.

That is where things start getting interesting.

More responsibility is no longer the same thing as moving up

Gartner found that 75 percent of HR leaders said managers were overwhelmed by the expansion of their responsibilities. Gartner now describes managers as carrying 51 percent more responsibility than they can effectively handle.

And look at the list of things managers are increasingly expected to do.

Coach people.

Develop talent.

Manage performance.

Create psychological safety.

Execute strategy.

Lead change.

Coordinate across functions.

Navigate conflict.

Retain employees.

Interpret data.

Redesign workflows.

Figure out where AI belongs.

Supervise what AI produces.

Maintain culture.

And, depending on the organization, still do a considerable amount of the actual work themselves.

This isn’t simply a bigger management job.

It is several organizational jobs being compressed into one title.

That distinction matters because organizations can look at a manager’s increased responsibility and interpret it as increased leadership. The manager experiences something very different: more surfaces on which they can fail without a corresponding increase in their ability to determine what happens.

Which creates a peculiar kind of organizational inflation.

You can receive the title, salary bump, additional reports, and strategic expectations traditionally associated with moving upward while discovering that the actual purchasing power of your authority has barely changed.

You are richer in responsibility.

You may not be richer in power.

COHERENCE

The managerial version of inflation

Imagine two directors.

Both have the same title.

Both manage teams.

Both sit in leadership meetings.

Both are told that they “own” important outcomes.

Director A controls a budget, can change priorities, has discretion over staffing, has direct access to the executive making the decision, and can kill work that no longer makes sense.

Director B is accountable for the exact same outcome but needs approval to change headcount, approval to move money, approval to change the deadline, alignment from three other departments, and executive sign-off to abandon a priority everyone privately agrees is failing.

On the org chart, these people are peers.

Operationally, they do not have the same job.

Responsibility is becoming a terrible proxy for authority.

That creates a problem because many of our assumptions about leadership still treat the two as though they rise together. We tell people that greater responsibility means greater influence when, in many modern organizations, responsibility can expand much faster than decision rights.

This is where a lot of supposedly individual leadership problems begin looking suspiciously structural.

The manager who “struggles to prioritize” may have eight priorities because she lacks the authority to eliminate three of them.

The manager who “needs to delegate more” may be working in a system where every meaningful decision still escalates back to her.

The director who “needs to be more strategic” may spend 70 percent of his week resolving operational problems created by a span of control the organization doubled in the name of efficiency.

The person who “needs to lead change” may have been handed accountability for implementation six months after someone else made every consequential decision.

We diagnose the person.

Then we send them to leadership development.

Gartner’s research is particularly interesting here because it argues that more training alone is unlikely to solve the problem. If the job itself has become unmanageable, increasing the manager’s skill without redesigning the job mostly produces a more skilled person doing an impossible amount of work.

That is like responding to inflation by teaching people to become better shoppers.

Helpful at the margins.

Not really the problem.

The job is starting to split

There is an even more literal sign that something is happening to the category.

Telstra, the Australian telecommunications company, separated the traditional manager role into two jobs: a leader of people and a leader of work.

The people leader focuses on capability, development, and helping employees move through the organization. The work leader focuses on getting projects delivered, without directly managing employees or controlling budgets.

I find this fascinating because it makes explicit something many organizations are still pretending isn’t happening.

“Manager” may have been a bundle.

People leadership, work coordination, technical judgment, resource allocation, performance management, organizational translation, and decision authority were packaged together because the industrial organization made that combination useful.

That does not mean they must remain packaged together forever.

Deloitte describes a life-sciences company that experimented with nearly tripling managers’ spans of control. The managers could keep the operation running, but coaching and employee development deteriorated, engagement declined, and turnover increased.

That experiment tells us something important.

The administrative capacity to supervise more people is not the same thing as the human capacity to manage them well.

Technology can make it possible for one manager to approve 30 timecards, monitor 30 dashboards, distribute information to 30 employees, and receive performance data about all of them.

None of that means the manager suddenly has three times the judgment, attention, emotional capacity, coaching time, or contextual understanding.

The spreadsheet scales.

The human doesn’t.

And then AI arrives

This becomes even stranger as AI enters management.

The early promise was straightforward: automate administrative work and managers will finally have more time to lead.

Some of that will happen.

But AI is also creating an entirely new category of managerial responsibility.

Gartner now argues that managers will increasingly be accountable not only for people and work, but for the behavior and outputs of autonomous digital agents. Someone still has to verify what the system was asked to do, evaluate whether its output makes sense, recognize exceptions, determine when human judgment is required, and own the consequences when something goes wrong.

Which means AI may remove tasks without necessarily removing accountability.

That distinction should make every executive interested in AI pause for a second.

Because organizations have a habit of counting what technology eliminates more carefully than they count what technology creates.

We remove ten hours of administrative work.

Great.

Then we add responsibility for supervising AI output, redesigning workflows, retraining the team, handling exceptions, reassuring employees, maintaining quality, governing risk, and figuring out what the new operating model is supposed to be.

Congratulations.

We saved the manager ten hours.

We also gave her a small consulting practice.

ACTION

Stop asking whether managers can handle more

The better question is what they can actually control.

If you lead an organization, take one consequential management role and list everything that person is accountable for.

Then, next to each responsibility, ask:

What decision can this person make without asking someone else?

The gap may surprise you.

If a manager owns retention but cannot change compensation, staffing, workload, hiring criteria, schedules, promotion decisions, or team structure, what does “owns retention” actually mean?

If a director owns execution but cannot remove a priority, move resources, change a deadline, or resolve a cross-functional dependency, what exactly are we holding them accountable for?

This isn’t an argument that every manager should have unlimited autonomy.

Organizations need constraints. Some decisions should remain centralized. Coordination matters.

But the constraint needs to be visible.

You cannot responsibly assign accountability without examining the authority required to carry it. (Which, by the way, is the work I do.)

Otherwise organizations create jobs in which people are close enough to the outcome to be blamed for it but too far from the decision to materially change it.

And once that happens, predictable behaviors emerge.

More meetings.

More escalation.

More stakeholder management.

More documentation.

More consensus seeking.

More requests for alignment.

More risk aversion.

None of those necessarily mean the organization hired timid leaders.

They may be perfectly rational adaptations to a system in which the cost of making the wrong decision remains with you while the authority to make the right one sits somewhere else.

Maybe we need to stop treating management as one job

This might ultimately be the larger implication.

Instead of asking how to make managers better at carrying an ever-expanding bundle of responsibilities, organizations may need to decide which responsibilities actually belong together.

Who develops people?

Who coordinates work?

Who owns technical judgment?

Who allocates resources?

Who resolves competing priorities?

Who has the authority to stop something?

Who translates evidence into a decision?

Those responsibilities can live in the same person.

But they should live there intentionally.

Telstra’s experiment is interesting precisely because it treats the traditional management bundle as a design choice rather than a law of nature.

That is probably where more organizations are headed.

Not toward a world without management.

Toward a world in which we finally admit that “manager” has been hiding several different kinds of work.

SIGNAL BOOST

Watch the distance between responsibility and authority

For years, we have measured leadership progression partly by looking upward.

How many people do you manage?

How big is your budget?

What is your title?

How many functions report to you?

Those measures aren’t useless.

But I suspect a more important measure is emerging:

What can you actually decide?

Two people can have identical titles, identical salaries, and identical responsibility for an outcome while possessing radically different abilities to influence it.

The person with 40 direct reports may have less usable authority than the person with four.

The VP may have an impressive title and spend half the week obtaining permission.

The manager may technically “own” the result while operating inside a system where nearly every lever affecting that result belongs to someone else.

And that is why I don’t think the interesting story is that middle managers are disappearing.

Something stranger is happening.

The category still exists. The bundle underneath it is coming apart.

Organizations are redistributing the component pieces of management at different speeds. Responsibility is moving. Authority is moving. Work is moving. Technology is changing what requires human judgment. Hierarchies are flattening in some places while decision-making remains stubbornly centralized.

So we continue using the familiar language.

Manager.

Director.

Vice President.

Leader.

But the organizational reality underneath those words can now vary enormously.

That should matter to anyone trying to understand why execution keeps breaking down.

Because eventually you arrive at a person who knows what is happening, understands what should probably happen next, and is accountable for what happens if it doesn’t.

Then you discover they cannot actually decide.

And suddenly a lot of what organizations call a leadership problem looks like something else entirely.

It is an authority problem wearing an accountability costume.

For the first five years of my professional career, I believed a fairy tale.

I thought the hard part was finding the right answer. I spent years helping leaders make sense of messy evidence, turning the data over, adding the qualitative context, and looking for the thing everybody else had missed.

I assumed that once we’d found the best insight, the rest would take care of itself.

Then something peculiar happened.

We found the insight. Everyone understood it. And nobody did anything.

Then it happened again. Except the second time, the organization eventually did the opposite of what the insight recommended.

Yes. The opposite.

Then it happened again, and then again. Eventually it stopped looking like an anomaly and started looking like the pattern. The anomaly was when you could actually draw a clean, logical line between what the evidence suggested and what the organization did.

The usual pattern looked different. Analyze the evidence. Find the thing that matters. Present it clearly. Then watch it get discussed, softened, delayed, half-used, or changed until it couldn’t threaten anything important.

Sometimes the very people who’d asked for the analysis got angry when the analysis contradicted the story they’d already built. Then came the familiar chorus: We need more data. We need more stakeholder input. We need more time.

Meanwhile, the thing everyone claimed to need was sitting right there.

You asked for it. We found it. You’re not using it.

That frustration changed my work. I stopped studying only how organizations find good insight and started paying much closer attention to what happens after the insight is found. I wanted to understand why smart people could see something clearly and still struggle to act on it.

Finding what’s true is not the finish line.

Sometimes it’s where the trouble starts. Because the moment an insight becomes consequential, the organization has to deal not only with whether it’s true, but with what becomes necessary if everyone agrees that it is.

CLARITY

The problem is rarely the insight.

Smart organizations usually aren’t short on information. They’ve got dashboards, surveys, assessments, talented people, research teams, and consultants with tasteful slide decks and shoes that say, “I have airport status.”

The insight exists.

What organizations often lack is a way to survive what the insight implies. Because good information doesn’t enter an organization as information. It enters as implication.

If the evidence is true, something else may have to change.

A strategy may need to shift. A leader may have to walk back a commitment. A team may need to admit that the metric they trusted was measuring the wrong thing.

A department may have to stop defending a model around which it’s built budget, language, identity, and an alarming number of PowerPoint templates. Someone may have to acknowledge that the thing they confidently defended last quarter isn’t the thing the organization needs now.

That’s why the gap between knowing and acting usually isn’t a gap in intelligence.

It is a cost problem.

Accepting good insight is expensive. It’s not always financially expensive, either. More often it’s socially expensive, politically expensive, strategically expensive, or expensive to someone’s identity.

It’s expensive in the way that requires someone to sit in a meeting and say, with their whole face, “The thing I defended last quarter may not be the thing we need now.” Until someone understands that cost, even accurate, carefully researched evidence can find a way to stall.

Insight creates risk.

The simplest version is this: data insight tells you what the evidence reveals. People insight tells you what the evidence threatens.

You need both. Without data insight, organizations drift into opinion, instinct, politics, and noise. Without people insight, smart people make the very reasonable mistake of assuming the facts should be enough.

They are not.

Clear evidence can still die in the room if nobody understands what accepting it would cost. Being right tells you what should happen. It doesn’t tell you whether the organization can survive making it happen.

COHERENCE

I once watched a program nearly get killed by the very evaluation process designed to determine whether it was working.

A team had built something genuinely new. The design was strong, the early results were promising, and the work had the faint smell of something that might actually matter.

Then they evaluated it.

They used the same criteria they’d always used. The old rubric. The rubric built for the very model this program had been designed to move beyond.

By those standards, the program looked mediocre. By criteria that actually reflected what the program was trying to accomplish, it was exceptional.

The program was succeeding at exactly the thing everyone had claimed they wanted.

And it nearly died anyway. The issue wasn’t the evidence. The issue was what accepting the evidence would require of the people in the room.

Three different kinds of risk were operating at the same time.

1. Logic Risk

Logic risk appears when new evidence threatens the framework an organization’s been using to understand its work.

This is bigger than changing your mind about one initiative. Organizations build metrics, roles, structures, investments, reporting lines, meetings, and solemn little strategic phrases around their assumptions.

Once enough things have been built around a particular logic, that logic stops behaving like a hypothesis.

It starts behaving like furniture.

Nobody remembers exactly why the giant cabinet is in the dining room anymore, but several important people become strangely emotional when you suggest moving it. Bad coherence can become self-protective because careers, budgets, decisions, and reputations have accumulated around it.

In this case, accepting the new evaluation criteria meant acknowledging that the old criteria might not be trustworthy anymore.

That wasn’t a small methodological adjustment. It meant questioning decisions that’d already been built on top of the old logic. People rarely volunteer to saw through the floor they’re standing on.

2. Strategic Risk

Strategic risk appears when new evidence requires someone to unwind a commitment that’s already in motion.

Organizations announce things. Leaders champion initiatives. Budgets get approved, names get attached, and somebody eventually makes a kickoff speech involving the phrase, “This is who we are becoming.”

Once that happens, evidence pointing in another direction has a problem. Accepting it may require someone to absorb the cost of reversing a position publicly.

That is not a data problem.

It’s a reversal problem. And reversals become particularly expensive when the person being asked to make one has already attached their credibility to the original direction.

In the program I’m describing, prominent people had already committed to how the work would be assessed. Changing the criteria wasn’t only an evaluation decision. Someone would have to acknowledge that the standard they’d backed was the wrong standard.

Even in healthy organizations, that’s difficult. In unhealthy organizations, you might as well ask someone to walk barefoot across the conference table while Finance takes notes.

So organizations reach for softer language. We need more time. Let’s revisit the assumptions. We should socialize this with a few more stakeholders.

Sometimes those requests are legitimate.

And sometimes they mean: Please give my ego a medically supervised exit ramp. Distinguishing between those two possibilities is part of the work.

3. Social Risk

Social risk appears when accepting the evidence changes someone’s status in the room.

Organizations are social systems. Credibility, reputation, belonging, authority, and status are real currencies whether serious people like admitting it or not.

If accepting new evidence makes someone look incompetent, outdated, complicit, or simply wrong in front of people whose opinions matter, the social cost can become larger than the analytical question. Nobody has to say, “You were wrong.”

The room usually has more sophisticated technology than that.

A pause. A glance. Someone says “interesting” in a tone that means absolutely not. The room develops weather.

When logic, strategic, and social risk activate at the same time, organizations become remarkably creative. The methodology gets questioned. The timeline gets extended. A committee forms.

The committee creates a working group. The working group schedules a listening session. Eventually a summary document appears in a shared drive where it’ll remain untouched until the sun consumes the earth.

This is how insight dies.

Administratively. With calendar invites. Usually without anybody needing to explicitly say no.

The program I’m describing survived.

Barely.

It survived because someone finally named what the argument was actually about. The disagreement wasn’t really about the criteria. It was about the cost of changing them.

Once those costs became visible, the room could deal with them. The obvious work had been evaluation. The real work was helping the room survive what the evaluation revealed.

ACTION

You don’t make insight move by making it more correct.

That sounds obvious until you watch very smart people respond to resistance by adding more evidence to a room that already understands the evidence. Another chart appears. Then a longer appendix. Then a cleaner summary.

Eventually someone sends a pre-read with “for additional context” in the subject line, which is usually where hope goes to stretch its hamstrings immediately before pulling something.

Sometimes the people in the room don’t need more explanation.

Sometimes they understand perfectly.

That’s the problem. Because understanding the insight also means understanding what accepting it’ll require.

So once the evidence is solid, the question can’t only be, “Why are they resisting?”

Ask a different question: What will this insight cost, and who’ll be asked to pay?

That question changes the work. If the insight breaks the framework people have been using to understand the organization, you’re dealing with logic risk.

If it disrupts a plan, investment, timeline, or public commitment already in motion, you’re dealing with strategic risk. If accepting it makes someone look wrong, exposed, out of touch, or less competent, you’re dealing with social risk.

There’s another distinction worth making here because not all resistance deserves the same response.

Some resistance is intelligence. People aren’t rejecting the evidence. They’re warning you that the proposed path will fail.

Some resistance is risk. People understand the evidence but they’re struggling to absorb its consequences. And some resistance is obstruction, where complexity, process, tone, timing, methodology, or endless requests for proof become camouflage for avoiding accountability or protecting power.

Those are not the same thing.

So they can’t be handled the same way. Treating all resistance as ignorance is how organizations end up communicating harder when the problem was never communication.

Sometimes the work is reframing the issue so a leader can update without unnecessary humiliation. Sometimes the insight is right but the implementation path needs to change.

Sometimes the work needs resources because unfunded priorities aren’t strategies. They’re vibes with deadlines.

Sometimes the evidence reveals broken trust and the intervention is repair. Sometimes process has become camouflage for avoidance and the intervention is confrontation. Sometimes the wisest move is a pilot, an escalation, or simply stopping because sunk-cost loyalty has become damage.

The goal is not to make the truth softer.

It’s to make the truth survivable enough to move. Because accuracy alone doesn’t create action, and an insight that people can’t afford to accept will stall no matter how true it is.

Think about the last obvious decision in your organization that somehow went nowhere. The initiative everybody knew needed to change. The problem people kept studying after the answer had become clear.

Think about the thing everybody could see but nobody could safely say.

Was the insight wrong, or was the cost of accepting it too high?

If it was the second one, there’s a path forward. Costs can be named, and named costs can be managed.

The work isn’t simply to find better insight.

It is to make important insight survivable enough to change the organization.

That changes what happens after the analysis is done. Instead of pushing the answer harder, you diagnose the environment the answer has to survive.

SIGNAL BOOST

There’s a distinction I keep returning to:

Data insight tells you what the evidence reveals. People insight tells you what the evidence threatens.

That distinction has changed how I think about strategy, assessment, organizational change, AI implementation, leadership feedback, and decision-making. If you only have data insight, you can be right and still ineffective.

If you only have people insight, you can become politically sophisticated and completely untethered from reality. The work is holding both at the same time.

What the evidence reveals. What the evidence threatens. And what it’ll take for the organization to survive accepting both.

Clarity Over Noise is for leaders working in that space between evidence and action.

I write about what happens when organizations learn something they can no longer afford to ignore, but acting on it creates political, strategic, or organizational risk. If that’s the kind of problem you’re trying to solve, subscribe.

Because finding the insight is only the beginning.

The harder question is whether the organization can survive what it now knows.

That’s the space I’m interested in. Not just whether we can uncover what’s true, but whether we can make what’s true survive pressure, politics, and practice.

Now that you understand it, how do you do it?

You probably already have an insight in mind.

Maybe it’s the recommendation that keeps getting “one more round of feedback.” Maybe everybody agrees with the analysis, but somehow the decision keeps moving two weeks to the right.

Or maybe someone keeps asking for more evidence and you’re starting to suspect that evidence isn’t the problem anymore.

This is where the framework becomes useful.

I built the Insight Risk Diagnostic for exactly that moment. Bring one real insight, recommendation, or decision that’s stuck, and the diagnostic will help you figure out what’s actually happening around it.

By the time you finish, you should be able to answer five questions:

What does the evidence threaten?

Who’s being asked to absorb the cost?

Is the resistance intelligence, risk, or obstruction?

Which kind of risk is dominating the room?

And what should you do next?

This isn’t a reflection exercise.

You should leave with a clearer explanation for why an important insight is stuck and a next move that fits the actual problem, instead of adding another chart to a room that already understands the chart.

Read more

CLARITY

Most innovation begins with somebody saying something that could easily make them look stupid.

It might be a half-formed thought, a question with no obvious business case, or an email that ends in “haha,” which is the written equivalent of placing a tiny pillow beneath your own fall. The idea arrives before the person sharing it knows whether it is useful, defensible, or mildly embarrassing.

Nobody schedules this moment.

There just won’t ever be a calendar invitation titled: 11:00 a.m. to 11:30 a.m. Produce Unexpected Value Through Interpersonal Warmth. Instead, someone walks a file down the hall rather than attaching it to an email. Someone stays after the meeting because another person looked unconvinced, or hears a weak idea and resists the ancient managerial urge to kill it before lunch.

These actions look minor because they are minor.

That is partly why they matter. Innovation often begins as a social favor before it becomes an organizational asset. You give someone a little more time than the task requires, ask one more question, or treat their unfinished thought as something worth examining rather than a small animal that wandered into the conference room and must be removed immediately.

That is kindness at work.

It is not kindness as decoration. It is kindness as infrastructure: the relational support that allows an unfinished idea to remain alive long enough to become more intelligent.

COHERENCE

Workplaces often confuse kindness with niceness.

Niceness keeps the room pleasant. It smooths the rough edges, avoids the awkward question, and sends the email that says “Great thoughts!” shortly before everybody ignores the thoughts forever.

Kindness is more demanding.

It allows someone to bring an idea into the room before the idea has acquired a blazer, a three-year forecast, and six slides explaining its strategic alignment. Niceness avoids tension. Kindness gives tension somewhere productive to go.

That distinction matters because unfinished ideas are embarrassing.

A finished idea can defend itself. It has evidence, language, and a clear account of the problem it solves. It may even have a logo.

An unfinished idea arrives wearing one sock.

It needs other people to help it discover what it is. Someone has to ask the question the originator has not yet considered, identify the part worth keeping, or explain why the entire thing is nonsense without making the person regret having spoken.

Daniel Kahneman described his collaboration with Amos Tversky as beginning, in part, because they enjoyed working together.

Before their ideas changed psychology, the relationship made it pleasurable to keep thinking in the same room. This sounds almost offensively simple: two brilliant people liked talking to each other. But intellectual history is full of ideas that emerged because somebody wanted to continue the conversation.

Airbnb has a similarly undignified origin story.

One of its founders sent his roommates a self-deprecating email suggesting they rent out their apartment. The message did not arrive with the confidence of a man unveiling a new category of global hospitality. It ended with “haha.”

That “haha” was doing a lot of work.

It said: I know this may be ridiculous. Please do not punish me for briefly imagining it. The idea could survive because the relationship could hold it before the market could.

That is what kindness does.

It lowers the social cost of being early. It gives people permission to share thoughts before those thoughts are polished enough to protect the person who produced them.

ACTION

Organizations say they want innovation, then design environments where nobody wants to be seen producing anything unfinished.

They ask for bold thinking in rooms where every sentence is evaluated for competence. They encourage experimentation but require the experimenter to explain the return before the experiment has taught them anything. They tell people to challenge assumptions, preferably after those assumptions have been approved by the steering committee.

Then leadership wonders why all the ideas feel strangely familiar.

The problem may not be creativity. People may have learned to reveal an idea only after it is polished enough to be politically harmless. By then, much of the interesting part has been sanded away.

A workplace that wants innovation needs to protect several small behaviors.

Someone has to walk the document over, ask what another person meant instead of deciding they were wrong, or let a strange idea breathe for five minutes before introducing it to Finance. Someone has to invite the person whose thought is difficult to summarize and continue the conversation after the meeting has performed its ceremonial death.

None of this guarantees innovation.

Kind people produce terrible ideas too. Sometimes the kindest possible response is, “I love you, but this proposal appears to have been assembled during a gas leak.” Kindness does not require pretending every idea is good. It requires giving an idea enough dignity to discover whether it is.

That is a different operating standard.

The question is not whether the idea was impressive when it arrived. The question is whether the way people treated each other allowed the idea to become more intelligent than it was at the beginning.

SIGNAL BOOST

You can learn a lot about an organization by watching what happens after someone says, “This may sound stupid.”

In some workplaces, that sentence opens a door. In others, it is a final warning before the speaker is professionally tranquilized. The difference reveals whether the organization merely praises innovation or has created the social conditions required for it.

New knowledge rarely enters an organization looking like knowledge.

It enters as a suspicion, an awkward question, a bad first draft, or an email ending in “haha.” At that stage, the idea cannot defend itself with evidence or authority. It survives because somebody chooses not to punish the person who brought it forward.

Later, if the idea becomes consequential, the organization may have to absorb the cost of knowing it.

The idea may challenge a strategy, expose a bad assumption, threaten someone’s status, or require the organization to change how work gets done. But before an insight can create that larger risk, someone has to make it safe enough to be known.

Innovation depends on talent, resources, incentives, timing, and luck.

But all of those ingredients still have to pass through relationships. People must share what they noticed, admit what they do not understand, and bring forward ideas that have not yet learned how to protect themselves.

Kindness makes that possible.

It is not the soft alternative to serious work. It is one of the conditions that allows people to think together long enough for serious work to become something nobody could have planned in advance.

An organization cannot act on ideas that its relationships do not allow people to reveal.

This issue is near and dear to my heart because it describes a hurdle I spent years trying to overcome in my own career.

What follows is a distillation of what I learned, what I wish somebody had told me sooner, and what I have since stress-tested with other people struggling with the same problem. My hope is that in the five or ten minutes it takes you to read this, I can save you five or ten months, maybe even a few years, of running very efficiently on a treadmill pointed at the wrong destination.

CLARITY

Being good at your job creates a strange problem.

People start sending you more of the work you are good at. You solve it faster than everyone else, which makes them trust you with even more of it. Eventually, the organization has built a very efficient machine for ensuring that you spend your career doing roughly the thing you were already good at.

This is usually presented as a compliment. You are the marketing person. The numbers person. The one who can rescue a difficult client, fix the presentation at 11 p.m., or somehow persuade Derek from Legal to answer an email before the heat death of the universe.

All of that feels good until you want to become something else.

Your competence can keep expanding while everyone’s idea of your competence stays exactly where it was. You can learn new skills, take on unfamiliar responsibilities, and solve problems far beyond your original job description. The people around you may see every bit of it and still fail to update their understanding of who you are.

I have started thinking of this as competence gravity.

The better known you become for one kind of contribution, the more everything you do gets pulled back toward it. Your reputation develops enough mass that new evidence does not necessarily change the story people tell about you. Sometimes the new evidence gets swallowed by the old story.

Say you are known as an exceptional marketing leader and you begin doing serious operational work. You redesign workflows, improve cross-functional coordination, fix broken processes, and become the person senior leaders call when execution gets messy.

You think you are demonstrating operations capability.

Your colleagues may be thinking, Wow, he is really good at marketing. They are seeing the same evidence you are. They are simply filing it in a different folder.

This is how competence becomes a trap.

You are producing evidence of a new professional identity, but the organization keeps using it to confirm the old one. Because the old identity is usually based on something you genuinely do well, every success can make that identity harder to escape.

We tend to assume careers are evidence-based. Do work at a higher level, people see that you can operate at a higher level, and eventually the title catches up. There is a wholesome little meritocratic story embedded in there, and I regret to inform you that organizations have once again found a way to make things weirder.

Evidence does not interpret itself.

People interpret evidence through whatever model of you they already have. Once that model has become useful to an organization, changing it can be much harder than developing the underlying capability in the first place.

COHERENCE

I was working with a client who was a VP of Marketing.

He was very good at marketing, which was part of the problem. His colleagues had years of evidence telling them exactly how to think about him. Marketing was no longer merely his function. It had become his organizational identity.

Earlier in his career, though, he had also done significant operations work. Over time, operations started creeping back into his current role.

Not fake operations either. He was not reorganizing a Monday.com board and putting “operational excellence” in his bio. He was solving actual operating problems: improving how work moved across functions, identifying bottlenecks, fixing handoffs, clarifying ownership, and taking responsibility for issues well outside a conventional marketing remit.

The work had changed.

The interpretation had not.

His colleagues could see that he was solving harder, broader problems, but the mental category they used to understand him remained remarkably stable.

When he did excellent operational work, they did not necessarily think, He has become a strong operator. Instead, the operational work became another example of why he was such an unusually effective marketing executive.

Which is an impressive career magic trick. Demonstrate competence in a second discipline and somehow strengthen your reputation in the first one.

This is where a lot of career advice becomes useless.

We tell people who want a different role to “show that you can do the work.” Take the stretch assignment. Volunteer for the cross-functional project. Start operating at the next level before you have the title.

That advice is not wrong. It is simply incomplete.

My client was doing the work. His problem was not a shortage of evidence. His problem was that the organization had already decided what the evidence meant.

Evidence only changes perception when people know what the evidence is evidence of.

Without that interpretive shift, people fit your new behavior into the story they already have. A finance leader starts shaping enterprise strategy and becomes known as someone who “brings a really strong financial lens to strategy.” An HR executive becomes central to an acquisition and gets praised for being “an incredible people partner.” A project manager starts developing junior leaders and becomes “so good at building project teams.”

The new competence gets colonized by the old reputation.

Nobody has to be malicious for this to happen. In fact, the phenomenon becomes more useful once we stop assuming somebody must be conspiring to keep us in our box.

Organizations need shortcuts because nobody has the cognitive budget to conduct a fresh epistemological review of Keisha from Procurement every Tuesday morning.

You saw Keisha yesterday. She was Procurement then. Barring an extremely eventful evening, she is probably Procurement today.

So we compress people because we have to. Sarah is numbers. Andre is partnerships. Miguel knows product. Keisha can explain why the vendor contract you were about to sign would result in everyone going to prison.

These shortcuts are incredibly useful until you need yours to change.

Then you discover that your professional reputation is not merely a summary of what you have done. It also determines what people notice when you do something new.

High performers can have an especially difficult time escaping because the original label is usually accurate. If everyone wrongly thinks you are bad at something, contradictory evidence has somewhere to go. But if everyone correctly thinks you are exceptional at something, you are asking them to loosen a story that keeps getting reinforced by reality.

Then a crisis hits.

The company has a major marketing problem, so who gets called? You solve it because, unfortunately, you continue to be annoyingly competent. Everyone feels relief, someone tells you how indispensable you are, and another brick gets added to the wall you have been trying to climb over.

The better you perform inside the old identity, the harder it can become to escape it.

That is the part people miss when they talk about career development as an endless accumulation of skills. Careers are not only built through capability. They are also built through interpretation, and interpretation has inertia.

THE ORGANIZATIONAL VERSION OF THE PROBLEM

So far, I have described competence gravity as something that happens to a person.

There is another side to it. Organizations can become so good at remembering what somebody has been that they become bad at seeing what that person has become.

That creates a talent problem.

The marketing leader who has become an excellent operator gets passed over when the COO role opens. The HR executive who has spent three years shaping enterprise strategy is still considered “the people person.” The director everyone already trusts with VP-sized problems somehow never appears on the VP succession slate.

Then the organization goes outside, pays a search firm, hires somebody with the “right” story, and occasionally loses the person who had been doing half the work already.

We tend to call this a retention problem after they resign.

It started much earlier as an interpretation problem. The organization had capability in the building. It simply could not see that capability clearly enough to deploy it.

This is why competence gravity is more than a career issue.

If your organization cannot update its understanding of people as quickly as people develop, you eventually start allocating opportunity based on outdated information. Promotions, succession plans, stretch assignments, and leadership pipelines begin reflecting who people used to be rather than what they can now do.

Talent you cannot accurately see is talent you cannot intelligently deploy.

For senior leaders, that should be mildly terrifying. You can have excellent people, excellent intentions, and a very polished talent review process, then still make bad talent decisions because the categories underneath the process are stale.

A nine-box grid cannot save you from a bad mental model.

YOU PROBABLY DO NOT NEED ANOTHER CERTIFICATE

When people discover they have been pigeonholed, their instinct is often to acquire more evidence.

Another certification. Another committee. Another stretch assignment. Another year of doing pieces of the job they actually want while continuing to perform the job everyone already associates with them.

At some point, this becomes professional Pokémon.

You keep collecting capabilities and hoping somebody eventually notices you have completed the set. But if the organization has already decided what kind of person you are, another badge may simply become another interesting feature of the same character.

More competence is not always the answer.

If the problem is perception, you have to intervene in perception. That means changing how the work is described, who describes it, where it happens, and what other people are asked to conclude from it.

You are not simply trying to prove that you can do something new.

You are trying to make the old explanation of you increasingly inadequate. That is a different job, and it requires much more than hoping your work eventually speaks for itself.

ACTION

So how do you change the story without immediately leaving?

Leaving is always available, at least in theory. Sometimes it is also the correct answer. But I am interested in the harder question: What can you do if the organization is worth staying in and you want to expand how people see you from inside it?

I would work on four things.

1. NAME THE NEW COMPETENCE CORRECTLY

Language creates categories.

If my client described his operational work as “improving marketing effectiveness,” he was accidentally reinforcing the very perception he wanted to escape. The work needed to be described according to what it actually demonstrated: cross-functional operating discipline, process redesign, resource allocation, execution, workflow design, and organizational coordination.

This is not about inflating your accomplishments. It is about refusing to translate every new capability back into the language of the role people already associate with you.

If you spent a year redesigning how four departments work together, do not describe that as “helping marketing collaborate more effectively.”

You redesigned how four departments work together.

Organizations build identities partly out of repetition. If every description of your contribution begins and ends inside your existing functional category, you are helping everyone preserve the existing story.

2. CHANGE WHO GETS TO NARRATE WHAT YOU CAN DO

Self-advocacy matters, but professional reputation becomes much more powerful when other people carry the story.

There is a large difference between telling the CEO, “I think I have developed strong operations capability,” and having the COO tell the CEO, “He has become one of the people I rely on when we have a cross-functional operating problem.”

The underlying work may be identical. The organizational meaning is not.

This is one reason sponsorship matters.

A mentor helps expand what you can do. A sponsor can expand what other people believe you are capable of doing. At their best, sponsors do more than advocate for promotion. They help rewrite the category the organization has assigned to you.

If you are the only person describing your transformation, people can treat it as aspiration. When credible people start describing it, it becomes reputation.

3. MAKE YOUR OLD JOB SURVIVE WITHOUT YOU

This one hurts.

If you want the organization to stop seeing you primarily as the marketing person, you eventually have to stop being the answer to every important marketing question. As long as the old function remains highly dependent on you, the organization has a perfectly rational reason to keep you attached to it.

Indispensability feels like power.

Sometimes it is a cage with very flattering performance reviews. The compliments keep coming, the bonus may even be nice, and somehow you are still doing the same work three years later.

Breaking that dependency may require developing someone underneath you, redesigning decision rights, documenting what previously lived in your head, or allowing someone else to take visible ownership of work you could probably still do better yourself.

And yes, they may do it slightly worse at first.

That is often the price of professional expansion. If the only way your old function succeeds is through your continued personal heroics, you have not outgrown the role yet. You have become fused to it.

4. FORCE THE FUTURE INTO THE CONVERSATION

Talented people often spend far too long hoping the organization will eventually notice something that already feels obvious to them.

Maybe it will. I would not build a career strategy around spontaneous corporate enlightenment.

At some point, you need to make the category change explicit.

For my client, the conversation could sound like this:

“I know my reputation here has largely been built around marketing. Over the last 18 months, a significant part of my contribution has expanded into operations and enterprise execution. I want to understand what evidence you would need to see to consider me for a broader operating role.”

That question converts a fuzzy question about identity into a testable question about evidence. Maybe they need P&L ownership. Maybe they need you to lead a major cross-functional transformation. Maybe they want broader team responsibility or ownership of a specific operating metric.

Good. Now you know.

Or maybe you receive seven minutes of executive word salad about “broadening your exposure,” “continuing the journey,” and “leaning into opportunities,” at the end of which you somehow know less about your career than you did before the meeting.

That is information too.

You are trying to discover whether the organization has an actual path for updating its model of you. If it does, you can work the path. If every answer remains permanently vague, you may be dealing with an organization that enjoys your expanded contribution considerably more than it enjoys the idea of expanding your role.

THE PART WHERE THIS SOMETIMES DOES NOT WORK

You can do all of this correctly and still fail.

Organizations are not infinitely editable. Sometimes your history inside a place is simply too strong. Sometimes the executive who hired you still sees the person you were six years ago. Sometimes the organization desperately needs the thing you are trying to stop being.

And sometimes something more troubling is happening.

An organization may trust you to perform beyond your role long before it is willing to recognize you beyond your role.

That distinction matters. A company may happily give you broader problems, greater responsibility, harder decisions, and more complex work while continuing to insist that none of this fundamentally changes what you are qualified to become.

That arrangement can go on for years if you let it.

You keep learning, producing, and becoming more valuable. Somehow, you remain exactly where everyone expects to find you. At some point, you have to ask whether the expanded responsibility is producing expanded opportunity or simply making you extraordinarily useful exactly where you already are.

Growth without recognition eventually stops being development.

It becomes extraction. The organization has figured out that you can operate beyond the boundaries of your role and has incorporated that fact into how it uses you, without incorporating it into how it sees your future.

That is why people occasionally receive a strange form of career validation when they leave.

They spend three years trying to convince one company they can operate at a broader level. Then another organization meets them, looks at the exact same body of work, and hires them into the broader role in three weeks.

Same person. Same competence. Same résumé.

Different interpretation.

The new company has one enormous advantage: It does not remember you.

There is no old version of you sitting in everybody’s head competing with the person standing in front of them. Nobody remembers when you were the junior person, the marketing person, the project person, or the person who used to need approval before making decisions you now make every day.

That is why leaving sometimes works so spectacularly.

But I would not leave before diagnosing the problem correctly. Changing jobs is a much more expensive intervention than changing the story around your work, and you want to know whether you have actually tried to change the system producing the perception.

You do not break out of a professional box by simply proving you are bigger than it.

You change the language around your work, the people who validate it, the dependencies that keep pulling you backward, and the conversation about what your next role requires. You make it increasingly difficult for the organization to explain your contribution using the old category alone.

If you do all of that and they still cannot see you differently, you have learned something important.

You do not have a competence problem. You have an interpretation problem.

At some point, continuing to accumulate evidence for people who have already decided what that evidence means is just another form of running on the treadmill.

The hardest part may be accepting when it is time to stop running.

THE LEADERSHIP QUESTION

There is a version of this problem for leaders, too.

Who in your organization are you still interpreting through a version of them that is three years out of date? Who keeps getting bigger problems without bigger authority? Who has become so valuable in their current category that nobody is seriously considering them for another one?

Those are not soft talent questions.

They shape succession, retention, internal mobility, leadership pipelines, and whether your best people see a future with you before somebody else sees one for them. They also shape whether you are allocating expensive opportunities to the people most capable of using them or simply to the people whose stories are easiest to understand.

A useful talent review should therefore ask a question most talent reviews do not:

Where has this person’s capability changed faster than our category for them?

That is a different question from “Are they high potential?” or “Are they ready now?” It forces leaders to examine whether the model itself is stale.

Organizations do not lose people only when compensation is wrong or managers are bad. Sometimes they lose people because the organization keeps recognizing an older version of them long after a newer version has arrived.

SIGNAL BOOST

Sociologists William Isaac Thomas and Dorothy Swaine Thomas gave us an idea that applies beautifully here: When people define a situation as real, it becomes real in its consequences.

Professional reputations work that way.

If an organization defines you as a marketing leader, that belief shapes which problems come your way, which rooms you enter, which successors get developed beneath you, which risks people trust you to take, and which future roles seem natural for you.

Over time, perception starts generating its own evidence.

Your reputation begins as a description of your career. Eventually, if you are not careful, it starts designing it.

For leaders, the same thing is true at scale. The categories you use to understand people eventually become the architecture through which opportunity moves.

If the categories are stale, the system will be too.

Recognition is useful. Intervention is better.

For paid subscribers, I built the Competence Gravity Playbook: five diagnostic signs, a reputation audit, a category-shift exercise, a manager conversation, a 90-day perception test, and criteria for knowing when it is time to stop trying.

I do not want you to finish this issue thinking, Well, that explains a lot.

I want you to know what to do Monday.

Read more

I grew up in the New York metropolitan area. It’s a place where much of the country’s business, brains, and money is concentrated. So naturally, everyone’s always dressing to impress.

A suit doesn’t even feel like you’re really dressed up in that part of the country. It’s more like the basic casual attire of the working world.

Then I got a job in the Pacific Northwest.

The Pacific Northwest is everything the New York City area isn’t. It’s slower, with wide-open spaces and a completely different relationship to time. You ask someone what they did over the weekend where I’m from, and they’ll say they went to a museum, an art show, a concert, worked on their startup, or partied. Ask someone the same question in the Inland Northwest, and they’ll literally say, “I spent the weekend at the lake. We caught some fish and camped.”

So when I moved to the Pacific Northwest, I put on what I thought was normal business attire for my first day at my new job: my best suit.

It was an entry-level job, and based solely on my attire, everyone looked at me like, “Is he our new boss?”

No one wears suits in the Inland Northwest unless they’re going to a wedding, a funeral, or an actual business meeting where they’re going to pitch something. Even then, they might not.

In one part of the country, that suit signified, “I’m ready to work hard, and I’m a professional.” In another part, the exact same suit basically said, “I’m either your new boss or I’m overdressed.”

Unfortunately, it was the latter.

I was the same person in the same suit. But as I walked in and out of different rooms, unbeknownst to me, a lot changed. The clothes hadn’t changed. The context had. And because the context had changed, so had the meaning of what I was doing.

That’s what happens to a lot of people when they move into leadership.

Clarity

People often arrive in leadership wearing the same metaphorical suit that helped them succeed before.

They’ve usually been promoted because they were excellent at the work. They solved problems quickly, knew the systems, caught the mistakes, calmed the clients, finished the reports, and rescued projects when everyone else had reached the “perhaps we should schedule another meeting” stage of organizational collapse.

Their competence was visible, so the organization rewarded it with a leadership role. Then everyone acted surprised when they kept doing the work that had earned them the promotion.

Leadership is a career change in disguise.

The promotion may happen inside the same organization. You might still sit near the same people, work on the same products, and attend meetings in the same aggressively beige conference rooms. On the surface, very little appears to have changed.

Underneath, almost everything has.

The habits that once communicated competence can start communicating something else. Stepping in quickly may have looked like initiative before the promotion. After it, the same instinct can tell the team you don’t trust their judgment. Having every answer may once have made you invaluable, but in a leadership role it can make everyone around you dependent. Rescuing a project might’ve been heroic when you were an individual contributor. Repeatedly rescuing projects as the leader can keep the people around you from learning how to recover without you.

The behavior hasn’t necessarily changed.

The room has.

That’s why high performance can get someone into leadership without preparing them for it. Organizations often promote people based on evidence that they can succeed in the old role, then evaluate them using expectations that belong to an entirely different one.

Before the promotion, the work may have centered on producing. Now it centers on creating the conditions for production.

Previously, knowing the answer was a strength. In leadership, helping other people develop judgment becomes just as important.

You once proved you could carry responsibility. The new role asks you to decide which responsibilities you should carry and which ones you need to release.

That transition is harder than organizations usually admit because it isn’t only operational. It’s also personal.

A promotion can look like a continuation of your career when it’s actually asking you to rebuild your identity inside it.

Coherence

Picture a high performer who’s just been promoted to lead the team.

At first, everyone’s relieved. She knows the work, understands the systems, and already has a reputation as the person to call when something goes sideways, catches fire, or acquires the mysterious phrase “executive visibility.”

So she keeps operating the way she always has.

Every important project gets her attention. Major decisions still pass through her. Weak drafts come back heavily rewritten because fixing them feels faster than coaching the person who produced them. A slipping deadline pulls her directly into the work. When someone asks a question, the answer comes quickly because she has it. If a team member struggles, she takes over because the work needs to get done and, at least in the moment, takeover feels responsible.

For a while, this looks like excellent leadership. Projects move, problems get solved, and senior leaders feel reassured because she’s involved in everything.

The costs appear later.

People gradually stop making decisions without her. Strong employees become more cautious because they know their work will probably be changed. Less confident employees become even less confident because they’ve learned that their first answer rarely survives contact with the boss.

Meanwhile, the new leader is working longer hours and becoming increasingly frustrated. From her perspective, she’s carrying the team. From the team’s perspective, she’s become the team.

Nobody set out to create this problem. It emerged because the leader is still measuring herself using the rules of the previous role.

Did I solve the problem?

Did I catch the error?

Did I make sure the work got done?

Those questions still matter, but they’re incomplete now. Leadership also requires a different set of questions.

Did I help the team understand the problem?

Was there enough clarity for someone else to make the decision?

Did I step in because the situation required me, or because being needed still feels like competence?

That last question usually has some bite.

Leadership asks people to loosen their grip on the behaviors that once earned them status. Someone who built an identity around being the person who knows now has to become the person who helps a group know what to do. A leader who once proved worth through personal output may now do their best work through someone else’s judgment, confidence, and performance.

That’s a difficult exchange for the ego.

There’s no award ceremony for the meeting you didn’t dominate. Nobody sends an email saying, “Congratulations on resisting the urge to rewrite Kevin’s entire proposal. Your commemorative mug is on the way.”

Still, restraint is often part of the work.

Leadership isn’t a clean break from management. Good leaders still have to manage, and good managers lead all the time. The more useful distinction is that the center of gravity has shifted.

Management often asks, “How will we get this done?” Leadership has to widen the frame: Why are we doing this? What will it require from people? What are we teaching them through the way we’re doing it?

The responsibility is no longer only to produce the outcome. It’s to build the conditions under which people can produce strong outcomes without needing the leader in every room.

Action

This is where leadership development often collapses into three familiar instructions: delegate more, trust your team, and stop micromanaging.

All reasonable.

Also nearly useless without structure.

A leader can’t wake up on Monday and decide to “trust more” in the abstract. Trust has to be translated into decisions, boundaries, and consequences. The team needs to know which decisions belong to the leader, which ones should sit closer to the work, and what someone is allowed to get wrong while learning. They also need clarity about when to escalate and what good judgment looks like before mastery arrives.

Without that structure, everyone stays trapped in the same loop. The leader intervenes because the team hesitates, while the team hesitates because the leader keeps intervening. Eventually, the organization announces an empowerment initiative, hands out a worksheet, and hopes the structural problem will become frightened and leave.

It won’t.

The role has to be reset explicitly.

That may mean separating the decisions only the leader can make from the decisions they’ve kept out of habit. It may require defining what ownership actually includes rather than telling people to “take more ownership” and then correcting every choice they make. Senior leaders may also need to stop rewarding heroic rescue work while claiming they want a more capable, distributed team.

You can’t praise someone for being indispensable every Friday and then act confused when nobody else can function by Monday.

This is the work I help leaders and leadership teams do when a high performer has been promoted but the organization’s old definition of competence is still running the system.

The goal isn’t to make leaders less involved. It’s to make their involvement more useful.

Strong leaders will still solve problems. They’ll simply stop proving their worth by being the only person who can.

Signal Boost

John Wooden once said:

“When the opportunity comes, it’s too late to prepare.”

That applies to leadership more than organizations usually admit.

Leadership development shouldn’t begin on the first day someone receives the title. By then, the pressure has already arrived, expectations have shifted, and the person is standing in a new room wearing the same suit that worked perfectly in the old one.

The real preparation is helping them understand, before the promotion, that the meaning of competence is about to change.

For paid subscribers

The essay above names the problem: a promotion can look like a small change while altering almost everything about how your competence is interpreted.

The strength that once made you valuable can become the habit that makes your team dependent.

The instinct that once helped you move quickly can become the reason everyone else slows down.

The standard that made your work excellent can become a standard nobody else knows how to meet without handing the work back to you.

That’s why this week’s paid tool isn’t another list of leadership habits.

It’s The Promotion Reset, a practical diagnostic for identifying which parts of your old role are still running your new one, then redesigning how decisions, standards, and responsibility move through your team.

You can use it privately, with a new manager, or as the structure for a leadership-team conversation.

Free readers get the argument.

Paid subscribers get the operating reset.

Read more

I built something that could help save physical books before they’re stripped of their bindings, scanned to train AI, then discarded.

Yes, we need legislation and protest. But those interventions take time, and books are still being lost while we wait for them to work.

We can’t un-destroy those books later.

The Book Preserve is designed to save as many as possible while we wait for something more permanent. It’s not a replacement for legislation or collective action. It’s a way to keep books alive long enough for those solutions to arrive.

When a bookseller receives an order that seems suspicious, they can post the book or collection on The Book Preserve with the price and a limited rescue window. During that window, a library, archivist, collector, or ordinary book lover can purchase it for the same price before the original sale goes through.

Think of it as a last-chance shelter for books.

A shelter may announce that an animal has a limited amount of time to be adopted before it is euthanized. The Book Preserve creates a similar window between the moment a book appears to be at risk and the moment it is permanently destroyed.

The bookseller still gets the sale.

But someone who wants the book to survive gets a chance to step in first. The seller doesn’t have to choose between preservation and getting paid.

The site is live as an early pilot. Real listings, payments, and preservation workflows are still being built, and I’m looking for booksellers willing to test a listing, preservationists willing to help assess books, and potential rescuers willing to test the experience.

Take a look:

bookpreserve.clarityovernoise.co

If you want to help test the pilot, send me a message here. If someone in your network should see this, please restack it.

Cardi B owns a Ferrari Purosangue.

If you don’t know what that is, just know it’s my dream car. Cardi, however, doesn’t have a driver’s license and doesn’t know how to drive.

Now, she’s earned every cent she’s made. She got it out the mud. She has every right to own that car.

But the Purosangue is known for how it drives. You can be chauffeured in it, of course, but that’s not really why you buy one. You buy it for the engine, the handling, and the feeling of being behind the wheel.

Which means I don’t think she’ll ever get the full return on that particular investment until she drives it herself.

So clearly, she should just give it to me. But I digress.

Every organization that has purchased enterprise AI has purchased a Ferrari. And if you’re going to buy a $500,000 car, you should probably know whether you’re getting everything you paid for.

That’s the enterprise AI question now.

Purchase means the capability belongs to you. Conversion means the capability is changing the work. It’s no longer enough to prove that employees have access, logged in, completed training, or attended a responsible-use webinar containing seventy-four slides.

The question is whether the investment is becoming anything.

Clarity

The first era of enterprise AI was about access.

How many licenses did we buy? How many employees completed the training? Did Legal approve the policy? Did IT turn the platform on without accidentally giving the entire company access to someone’s performance review folder?

These were reasonable questions. They were also the questions of an organization standing at the starting line.

Many companies have now crossed it. They’ve purchased enterprise platforms, formed committees, written policies, hosted training sessions, launched pilots, and created enough internal excitement to power a regional airport.

The licenses exist. The webinars happened. Someone created an AI champions channel in Teams. It has seventeen members, three unanswered posts, and one person who keeps sharing forty-minute videos with the message, “Worth a watch.”

It’s time to stop, Bob.

Access is no longer the interesting part. The interesting part is what that access became. Which workflows got faster? Which decisions improved? Which teams reduced rework?

Where did the organization recover time, lower costs, improve quality, support revenue, or reduce risk? And where is it still paying for possibility?

That’s the difference between adoption and conversion.

A company can have thousands of active users and still have no defensible account of what changed. It can point to logins, prompts, training completion, and usage dashboards without being able to explain whether any of it shortened cycle time, improved judgment, released capacity, prevented waste, or made the business better.

It can prove that people opened the tool. It cannot prove that the organization improved because they did.

I’ve talked to leaders who assure me their people are saving time by using AI to generate reports. Then I ask how much time those same employees spend fact-checking the reports, correcting weak claims, and re-prompting the system until the final version is defensible.

That number matters more.

It tells you whether time was actually saved or merely moved from writing to verification. And even when time was saved, the next question is whether it was redeployed into work that made the employee more useful, thoughtful, or impactful.

Otherwise, the organization may not have created capacity at all. It may have simply changed where the labor hides.

That’s the measurement gap sitting underneath much of the enterprise AI conversation.

Leaders are being asked to make renewal, expansion, budget, vendor, and governance decisions using evidence designed to answer a much smaller question: Did people use it?

Usage matters. It’s simply not the same as return.

A gym can tell you that someone scanned their membership card eleven times last month. It cannot tell you whether they exercised, sat in the sauna, or spent forty minutes taking mirror pictures with one dumbbell positioned nearby for plausible deniability.

A lot of enterprise AI dashboards have the same problem. They document presence.

Leadership needs evidence of consequence.

Coherence

Imagine two organizations buying the same enterprise AI platform.

They negotiate similar contracts, purchase roughly the same number of licenses, launch within the same quarter, and provide similar training. Six months later, both can report adoption.

From the dashboard, they may look almost identical.

Inside the work, they are not. At the first organization, a finance team has shortened a monthly reporting process by two days. A program office is producing first drafts faster but spending almost the same amount of time correcting them.

The communications team has reduced outside agency spending. A compliance group has restricted several uses because no one has measured the actual exposure.

Managers are also using AI to summarize meetings that should not have existed in the first place, proving once again that technology can make bureaucracy faster without making it less bureaucratic.

At the second organization, usage is broad but shallow. Employees generate emails, revise documents, summarize notes, and experiment with prompts. The platform is active, but the business is mostly unchanged.

Both organizations can prove activity. Only one can begin to prove return.

That difference won’t remain academic for long. It’ll arrive in the CFO’s office as a renewal question, enter the boardroom as an investment question, reach the CIO as a vendor question, and land with the COO as a workflow question.

The CEO asks whether the investment is changing the business. The CFO asks whether the return can be defended. The COO asks which workflows improved, while the CIO asks what should expand, be redesigned, renegotiated, or stopped.

These sound like four different questions. They’re four versions of the same one:

What did our AI investment become?

The answer cannot be another general strategy deck with tasteful gradients and the word transformation printed large enough to suggest transformation has already occurred. It requires measurement.

Start with cost and licenses. What did the organization buy? What’s being used? What’s sitting idle? Which teams have access to tools that don’t fit their work?

Then examine adoption reality. Who’s using AI deeply, shallowly, sporadically, or invisibly? Platform data can show frequency, but it rarely reveals the importance or quality of the use.

Then move into workflows. Where did cycle time change? Where did quality improve? Where was rework reduced? Did employees recover meaningful capacity, or did they simply produce more material for someone else to review?

Then examine enablement. Which teams need better training? Which need redesigned processes? Which lack useful applications because nobody has translated a broad technology into the work they perform every Tuesday morning?

Then assess risk. Where is the organization exposed through data handling, compliance, vendor dependence, reputation, trust, or decision quality?

A use case can create efficiency and still be a terrible idea. Speed is not absolution.

Finally, determine what should happen next. What should expand? What should pause? Where should the organization retrain, consolidate, renegotiate, redesign, or stop spending?

That’s where an executive picture begins to emerge.

A capability score can show where AI is becoming operational strength. A conversion gap can show how much paid-for potential has not yet become measurable return. Latent value zones can identify teams and workflows where better enablement is most likely to unlock something real.

An executive decision map can turn those findings into choices.

The purpose isn’t to create more information. Executives already have information. Some of them have so much information they now need a weekly meeting to discuss which dashboard they should ignore.

The purpose is to produce evidence strong enough to support a decision.

Action

The first step isn’t to build a perfect enterprise-wide measurement system.

That’s how a useful question becomes a fourteen-month initiative with three workstreams, six subcommittees, and a SharePoint folder called FINAL AI ROI FRAMEWORK v7.

Start smaller.

Choose one enterprise AI platform and a few priority functions. Establish what the organization is paying for, how the tool is actually being used, where important workflows have changed, and where capability remains trapped.

The work should combine several forms of evidence.

Platform and license data reveal access and visible adoption. Workflow analysis shows what changed in practice. Interviews explain what the dashboards cannot see.

Cost, time, quality, risk, and capacity measures help build a defensible return case. Training sessions and office hours can also reveal whether the obstacle is knowledge, workflow design, policy, confidence, or simply the absence of a meaningful use case.

No single source can answer the whole question.

Usage data without workflow evidence overstates progress. Interviews without operating numbers produce anecdotes. Financial estimates without employee experience miss how the work is actually being performed.

Policy review without measurement tells the organization where the borders are, but not where the money is.

A credible snapshot brings those pieces together. It tells leadership where return is compounding, where it is stuck, where exposure is rising, and what should happen during the next ninety days.

The immediate output might include a capability score by function, a measured conversion gap, latent value zones, an executive decision map, a board-ready account of return and risk, and a ninety-day action plan.

But the deeper outcome is decision discipline.

The organization stops asking whether AI is “going well,” a question vague enough to be answered with almost any chart and enough confidence. It begins asking better questions.

Which investments are changing the work? Which are mostly changing how the work is described? Which teams are ready to scale?

Which need redesign before another dollar is spent? What evidence would justify renewal? And what evidence would justify stopping?

This is the work Michael Marcus and I are building through the Enterprise AI Value Snapshot.

Michael establishes what the data can prove. He builds the measurement spine, data logic, analytics model, and defensible evidence base.

I translate what that evidence means for leadership, operating choices, organizational behavior, and the decisions that follow.

One side without the other is incomplete.

Technical measurement without executive interpretation creates reports that are accurate and inert. Executive interpretation without technical rigor creates a compelling story that may collapse the moment the CFO asks where the number came from.

The work is to connect them. To find what is true and make it move.

Signal Boost

The next stage of enterprise AI won’t be won by the organization with the most licenses.

It’ll be won by the organization that can answer the harder question: What did those licenses become?

Access tells you who can use AI. Conversion tells you whether the investment is changing the work.

For paid subscribers

The article gives you the argument.

The paid section gives you the tool.

The AI Conversion Gap Test helps you determine whether one AI platform, function, or workflow is creating measurable business value, or simply creating more activity that looks like progress.

Read more

Close your eyes and pretend you’re a doctor.

Someone walks into the clinic with a cough. You happen to be walking by, and from across the room, you point at them and say, “That’s bronchitis. Classic case, actually. Get them some antibiotics.”

Then you keep going on your merry way.

Obviously, that would be ridiculous. A cough is a symptom, not a diagnosis. A doctor has to get closer, listen to the lungs, and understand what’s actually happening before prescribing a remedy.

If there’s a wheezing sound, it may indicate that air is being forced through a narrow passageway, which could point toward asthma. If there’s a crackling sound, it may suggest fluid or an infection, which could point toward something like pneumonia.

The symptom may look similar.

But the diagnosis may be completely different. That matters because the diagnosis determines the remedy, and the wrong remedy isn’t neutral. It can waste time, mask the real problem, or make the situation worse.

If someone has asthma, you may be talking about an inhaler. If someone has a bacterial infection, you may be talking about antibiotics. Both people walked in coughing, but treating them as if they have the same problem would be irresponsible.

The same thing happens at work.

A project is delayed. A team is disengaged. A decision is stuck. From across the room, those problems can look similar.

But one may require management.

Another may require leadership. There is a difference between the two, and the difference is in the questions they ask and keep coming back to.

Management asks:

How do we get this done?

Managers think about objectives, processes, deadlines, resources, and accountability. They’re trying to make sure the machine runs, preferably without smoke coming out of it or someone sending an email marked “urgent” at 4:57 p.m.

Leadership asks:

Why are we doing this, and how will this impact people?

Leaders have to think about meaning, trust, direction, relationships, and what the work is asking of the people doing it. Management centers objectives. Leadership centers people.

Those questions require different skill sets, perspectives, and approaches.

But regardless of whether your job title has manager or leader in it, you’re going to be asked to do both. Titles describe the role someone hired you into. They don’t always describe what the room needs from you on a particular Tuesday.

Your title may be Manager of This and That, but today the room is asking you to lead. Your title may be an obvious leadership title, like CEO, but today the room is calling for you to manage.

You don’t get to say, “Sorry, I’m a leader. I don’t manage.”

You also don’t get to say, “Sorry, I’m a manager. I don’t lead. Go talk to Bob down the hall for that.” Bob is busy. And, if we’re being honest, Bob probably doesn’t know what’s happening either.

The real skill is being able to identify what the room is asking of you today, in this situation and in this season, then apply the related skills.

You have to develop the ability to manage and lead. But you also have to develop an additional skill that makes both of them maximally effective.

The ability to diagnose which one the room needs from you now.

Coherence

Here’s where a lot of leaders and managers begin and, unfortunately, stay.

Whichever one they’re already good at, they lean into it, develop it, and use it regardless of the situation. Success turns a skill into an identity, and once a skill becomes part of your identity, it gets much harder to notice when it has stopped being useful.

If a young professional is really good at managing, they’ll receive compliments on their management style. They’ll get promoted because of their ability to manage. They might even get a big leadership title, like VP of Something Cool, because of their ability to manage.

Then management becomes their default.

If it stops working, they lean even further into management and call it determination not to give up. More meetings. More tracking. More pressure. More execution. At some point, someone creates a dashboard to track why the other dashboards aren’t working.

Someone who’s considered a natural leader early in their career can make the same mistake in reverse. They lean into vision, inspiration, and relationships at the expense of creating systems, following up, or making sure the work actually gets done.

Both make the same mistake.

They apply what they’re best at instead of diagnosing what the room is asking of them. Their strength becomes the answer before they’ve even understood the question.

Cathy Engelbert and a title change in disguise

A textbook case of this may be what has played out with WNBA commissioner Cathy Engelbert.

ESPN’s Alexa Philippou interviewed more than 20 owners, management executives, players, and other industry insiders. The picture that emerged was that, for as much business growth as Engelbert has overseen, many stakeholders do not believe she is the right person to navigate the demands the league now faces and what its people need.

That distinction matters.

The issue is not whether Engelbert helped grow the business. By most accounts, she did. The issue is whether the skill set that helped grow the league is the same skill set the league most needs from her now.

When Engelbert was appointed in 2019, she became the league’s first commissioner. The people who held the position before her had been called presidents. She came into the role with more than 30 years of business and operational experience, including serving as the CEO of Deloitte.

The assignment was clear.

Take an excellent product and grow it. You had world-class athletes, literally the best women basketball players in the world, playing in a league whose revenue, reach, and salaries didn’t come close to reflecting their talent or cultural value.

For comparison, when Caitlin Clark entered the WNBA in 2024, she reportedly signed an eight-year Nike contract worth up to $28 million. Her WNBA salary for her rookie season was approximately $76,000.

She entered the league as perhaps the most famous women’s basketball player on earth and made less from the league than many experienced plumbers.

Shout-out to the plumbers. No shade at all.

The question in front of Engelbert was essentially: How do we take this excellent product and grow it so that the business begins to reflect the fact that these are literally the best women basketball players on earth?

That has management written all over it.

More sponsors. Better contracts. More investment. Expansion. Infrastructure. Revenue. Execution. The league had a strong product. It needed a stronger business wrapped around it.

And by many measures, it worked.

The league’s visibility grew. Interest surged. New investment arrived. Engelbert helped negotiate collective bargaining agreements, guided the league through the pandemic, and oversaw a period of major commercial growth.

But success changed the assignment.

The league became larger, more visible, and more culturally significant. With that growth came different questions from players, owners, teams, and the public.

Questions about trust, communication, safety, accountability, accessibility, and relationships. Questions about what kind of league the WNBA wanted to become as its influence grew.

That was a title change in disguise.

The job was no longer only to manage the growth. The room was asking Engelbert to lead through it.

Sometimes the difference between managing and leading shows up in what looks like a scheduling decision.

On June 21, the WNBA staged a game to honor its beginning. More than 18,000 fans were there. Legends from the league’s inaugural game were in the building. Nneka Ogwumike hit a buzzer-beating three to win it.

Engelbert wasn’t there.

She was 3,000 miles away at the U.S. Open, where she presented a trophy to a caddie. Later that week, she also missed the Women’s Basketball Hall of Fame induction honoring Cheryl Reeve, Candace Parker, and Elena Delle Donne.

Missing an event doesn’t prove someone can’t lead. Calendars are messy, and obligations compete. But leadership is often read symbolically.

A manager can look at two events on a calendar and ask, “How do we get this done?”

A leader has to ask, “What do these events mean, and how will my choice impact people?” That is the difference between managing competing commitments and understanding what your presence communicates.

Recent reporting has raised concerns about Engelbert’s relationships with players, the consistency of her communication, and whether stakeholders trust her to guide the league through its next phase.

Communication. Trust. Guide.

Those are blinking arrows pointing toward leadership, not management.

The issue isn’t that the business is being managed poorly. It’s that a management-shaped solution is being forced onto a leadership-shaped problem.

For example, ESPN asked Engelbert whether she felt misunderstood. That’s an obvious leadership question. It’s about people, meaning, and how she is being experienced.

Her answer moved immediately back to management:

That’s the key.

When faced with a leadership problem, she answered with execution. She didn’t begin with people or why. She began with the thing she knew she could do: execute.

That doesn’t mean execution stopped mattering. The business still had to run. Contracts still had to be negotiated. Expansion still had to be managed.

But the room had changed.

The people criticizing her weren’t primarily saying, “The strategy isn’t being executed.” They were raising concerns about trust, accessibility, communication, protection, and relationships.

Those aren’t problems you solve by executing the strategy harder.

A person can execute a strategy exceptionally well and still struggle to lead the people who have to live inside it. You can hit every milestone on the project plan and still have half the room mentally drafting their resignation letter.

The strength becomes the trap

This is what makes the transition so difficult.

The thing you’re best at probably brought you here. Engelbert’s business and operational abilities were part of why she was chosen, and they were essential to the league’s growth.

The mistake wouldn’t be having those strengths.

The mistake would be assuming that because those strengths worked before, the room will always need more of them. Yesterday’s solution can become today’s blind spot while still looking like competence.

That’s how a strength becomes a trap.

A manager receives a leadership title and keeps managing because management is what has always produced results. A leader becomes responsible for a complex operation and keeps offering vision because vision is what people have always praised.

Both are standing in front of a coughing patient, prescribing the remedy they know best before listening to the lungs.

Action

The next time you walk into a meeting, don’t begin by asking what your title gives you permission to do.

Ask what the room actually needs. The title may tell you where to sit, but it won’t tell you what the situation requires.

Does the room need clarity around timelines, roles, resources, accountability, and execution?

Then manage.

Ask what needs to get done, who owns what, what resources are required, where the process is breaking down, what the deadline is, and how everyone will know the work is complete. These aren’t lesser questions. Without them, the big vision becomes an expensive group conversation.

Or is the room struggling with meaning, trust, direction, relationships, or the human implications of the work?

Then lead.

Ask why you’re doing this, how it will impact people, what people are experiencing that you may not see, what the decision communicates, whose trust will be required, and what kind of organization you’re becoming through the way you do the work.

Sometimes the same meeting will require both.

You may need to lead for the first 20 minutes so people understand why the work matters and how it will affect them. Then you may need to manage for the next 20 minutes so everyone knows who’s doing what by Friday.

The skill isn’t choosing whether you’re a leader or a manager once and for all.

The skill is noticing when the room changes. A project may be delayed, people may be disengaged, a decision may be stuck, or a team may be underperforming. Those are symptoms, not diagnoses.

The delay could be a process problem that needs management. Or it could be a trust problem that needs leadership.

The disengagement could come from unclear responsibilities. Or it could come from people no longer understanding why the work matters.

The stuck decision could need more information. Or everyone may already know what needs to happen and simply be afraid of what telling the truth will cost them.

Applying management to a leadership problem often produces more tracking, more meetings, more dashboards, and more pressure.

Applying leadership to a management problem often produces an inspiring conversation followed by nobody knowing what they’re supposed to do next. Everyone leaves feeling transformed and then spends the afternoon asking, “Wait, who was supposed to send the spreadsheet?”

Management isn’t superior to leadership.

Leadership isn’t superior to management. They’re different remedies, designed for different conditions.

And the diagnosis determines which one will help.

Signal Boost

One team executive summarized the concern surrounding Engelbert this way:

“I do think Cathy thinks, well, she’s done her job. She got us financially set up the right way. It’s the other pieces that we’re talking about.”

It’s the other pieces.

The relationships, the trust, the communication, and the people. The parts of organizational life that don’t always fit neatly into a spreadsheet but eventually determine whether the spreadsheet means anything.

You may have done the job the room needed from you five years ago exceptionally well.

That success doesn’t guarantee you’ve noticed what the room is asking from you now. Sometimes the reward for solving one problem is being handed a completely different one while everyone keeps calling it the same job.

—

I’ve given you my best thinking in the free article.

I don’t believe in holding back the real idea and putting the good part behind the paywall. If this was all you needed, take it, use it, and shoot me an email to let me know how it’s going. Those messages are genuinely helpful and encouraging to me.

The paid section isn’t where I reveal what I really think.

It’s where I give you the tool I use to turn the idea into judgment.

A delayed project, a disengaged team, or a decision that keeps coming back from the dead may require management, leadership, or both. The hard part is diagnosing which one the room needs before you start prescribing your favorite remedy.

For paid subscribers, I created The Room Diagnosis, a practical 10-minute tool to help you:

• separate symptoms from diagnoses

• run the Management Test and Leadership Test

• identify mixed problems that require both

• choose the first intervention

• check whether you’re responding to the room or simply leaning into what you already do well

Because noticing the problem isn’t enough.

The diagnosis determines the remedy.

Read more

Have you ever heard of Maxwell Perkins?

Probably not.

But you’ve heard of F. Scott Fitzgerald, Ernest Hemingway, and Thomas Wolfe. Perkins edited all three at Scribner, but he didn’t edit them the same way. Fitzgerald needed encouragement and structural judgment, Hemingway needed restraint and trust, and Wolfe handed him enormous manuscripts that required aggressive shaping.

Same editor. Three completely different kinds of help.

That’s because good editing has never been about applying one formula to every writer. It’s about recognizing what the writer is trying to make, then helping them make it more fully. Sometimes that requires encouragement, sometimes it requires cutting, and sometimes it requires knowing when to leave the writing alone.

You’ve probably heard of T. S. Eliot too.

The version of The Waste Land we know today was shaped heavily by Ezra Pound, who cut long passages from a manuscript that was originally much larger. Eliot wrote the poem, but Pound helped him see what the poem needed to become. Eliot later called him il miglior fabbro, “the better craftsman.”

Under the logic Substack now seems to be using, some of the most celebrated writers of the twentieth century might’ve received a failing grade.

Substack has introduced a feature that scans writing for AI-generated or AI-assisted text. The problem isn’t only that tools like this can misidentify a person’s work based on some arbitrary definition of what “sounds like AI.” The deeper problem is that the entire idea misunderstands how writing works.

Substack is confusing authorship with solitude.

Many of the writers we now treat as singular geniuses didn’t produce their greatest work alone. They wrote in conversation with editors who questioned weak logic, cut repetition, challenged structure, suggested stronger framing, and identified things the writer was too close to the manuscript to see. The work belonged to the writer, but the process included other minds.

They didn’t call an editor and say, “Write me a novel.”

They wrote the novel. Then they asked another intelligence to help them make it better. The writer still brought the ideas, experiences, judgment, language, and final creative control.

That distinction matters.

There’s a meaningful difference between asking AI to generate a complete article you barely thought about and using it to challenge the framing of something you’ve spent months thinking through. There’s a difference between outsourcing your point of view and asking for help expressing it more clearly. There’s a difference between replacing the writer and editing the writing.

Substack’s grading system risks flattening all of those choices into a single accusation: AI touched this, so the work must be less authentic.

But writing has always been touched by other people. Editors have rewritten sentences, reorganized chapters, removed beloved passages, questioned arguments, and occasionally pushed manuscripts toward their own intellectual or political preferences. Some editorial relationships were so intrusive that they’d make today’s AI-assisted editing look almost timid.

No one concluded that the existence of an editor meant the author didn’t write the book.

We understood that authorship wasn’t determined by whether another person suggested a better structure. It was determined by where the thinking, judgment, experience, and point of view came from. The editor could shape the expression without becoming the source.

That’s still the right standard.

When I use AI as an editor, I’m not asking it to become me. I bring the experience, the argument, and the strange line that only makes sense because I’ve spent years sitting in meetings, studying organizations, and watching what pressure does to people. Then I might ask AI where the argument drifts, where the story loses energy, or whether the thing I’m trying to say is actually clear.

Sometimes it offers a better structure.

Sometimes it gives me a sentence I wouldn’t use in a million years. Sometimes it points toward something useful, and I rewrite it until it sounds like me again. Sometimes its best contribution is showing me exactly what I don’t want to say.

That isn’t the absence of authorship.

That is editing. The judgment still belongs to the writer, because the writer decides what stays, what goes, and what the piece ultimately means. Assistance doesn’t eliminate authorship unless the writer has surrendered the thinking itself.

Of course, AI raises real questions.

Writers should think carefully about what they outsource, what they disclose, and whether the technology is helping them clarify their thinking or avoid doing any thinking at all. Those are worthwhile questions because they deal with process, judgment, and integrity. But a detector can’t answer them.

It can’t tell the difference between someone who typed, “Write me a profound essay about leadership,” and someone who spent a year developing an idea, wrote the first draft themselves, and used AI to test the narrative arc.

It can’t see the notebooks, the abandoned drafts, the lived experience, or the dozen revisions that came before the final sentence. It can’t tell who made the creative decisions. It can only make guesses about the finished text.

And when those guesses are presented as grades, they turn a complicated question about process and authorship into a cheap verdict.

That’s especially strange on a platform built for writers. Substack should be helping writers think more deeply about authorship, attribution, and creative process. Instead, this feature risks teaching them to fear any sentence polished enough to trigger a machine’s suspicion.

Worse, it may reward writers for performing messiness so their work appears sufficiently human.

Misspell a word. Leave the structure loose. Keep the paragraph that doesn’t quite work. Apparently, imperfection is now proof of life.

I’m having none of it.

The question isn’t whether another intelligence touched the work. The question is whether the writer brought something worth touching. Did the idea come from lived experience, serious thought, original judgment, or genuine curiosity?

Did the writer make the final decisions?

Does the piece express a point of view that belongs to them? Can they explain why each argument, story, and sentence is there? Would the work still have a reason to exist without the tool?

If the answer is yes, assistance hasn’t erased authorship.

It has helped authorship become visible. A good editor doesn’t remove the writer from the work. A good editor helps the writer become more unmistakably themselves.

Maxwell Perkins didn’t make Hemingway less Hemingway.

Ezra Pound didn’t make Eliot less Eliot. Their involvement didn’t prove the writers lacked talent or originality. It proved that serious writing has rarely required the performance of absolute isolation.

AI can serve that same editorial function.

It can challenge a structure, reveal a gap, flag repetition, or suggest a framing the writer hadn’t considered. Whether that produces better writing or empty, polished slop depends on the judgment of the person using it. The presence of a tool can’t answer that question.

Substack should learn the difference.

By the time my dad graduated from medical school at the University of Ife in Nigeria, my mom had already built a successful career in operations.

Normally, their story would’ve kept moving upward. My dad would’ve gone into residency and begun practicing as Dr. Ayeni. My mom would’ve kept advancing in her career.

Instead, they immigrated to America.

Because my dad hadn’t yet completed residency, he couldn’t practice medicine in America yet. My mom’s experience didn’t immediately translate into the same kind of career she’d built in Nigeria either. Both of their careers effectively started over.

So they took whatever work they could find.

Janitorial work. Fast-food restaurants. Anything that’d keep their young family afloat. They worked those jobs right beside each other.

We were poor.

I didn’t fully understand that at the time. When you’re young, and your parents are doing everything they can to stand between you and the rain of the world, you don’t always realize there’s a storm. But looking back, I remember the cockroaches, the mice, and all the usual evidence.

Put any group of people under enough emotional, psychological, and financial pressure, and one of two things tends to happen.

They’ll turn on each other, and every person starts looking out for themselves. Or they’ll decide that nobody gets left behind. My parents had plenty of opportunities to choose the first option, but they repeatedly chose the second.

They chose to put their faith in God.

And they kept choosing each other.

They kept operating as a team. They kept having each other’s backs. They kept defining success the same way: We are all going to climb this mountain, or none of us will.

No other options.

Within a few years, my dad completed his residency and began working as a physician. My mom completed her degree and became a social worker. In the span of a few years, we’d gone from barely making ends meet to owning a home and no longer worrying about where the next meal would come from.

People would probably call that a miracle.

I do too. But having a front-row seat to it left me with two ideas I’ve never been able to shake. The first was the transformative power of education.

The second was even more important.

What becomes possible when people triple down on having each other’s backs.

Years later, when I was deciding what to study for my doctoral degree, I kept returning to that question. What if you could take what my parents did, bottle it, and reproduce it? What if teams and organizations could learn how to hold together under pressure instead of coming apart?

You could change more than a workplace.

You could change lives. So I began studying what that looks like inside organizations and teams. What I found changed my understanding of leadership.

I used to think leaders were the people who told others what to do and carried themselves with confidence. Leaders do those things. But those things are often the byproducts of leadership, not the heart of it.

The heart of leadership is building community.

The best leaders do that in different ways, but they all begin in the same place. They start with what is true. That became the next question that pulled me in.

In most organizations, the truth isn’t always hard to find. People usually know what isn’t working. They know which meeting is performative, which process is broken, and which idea died because the wrong person said it.

The problem is what happens next.

The truth gets softened, reworded, delayed, or buried. By the time it moves through pressure, politics, and practice, it often no longer resembles the thing everyone knew at the beginning.

That’s where my work began to come into focus.

I wanted to understand how to identify what’s true and turn it into genuine insight, help that insight survive pressure, politics, and practice, and use that truth as the foundation for building real community.

Today, when I work with leaders, teams, and organizations, I’m still doing what I watched my parents do all those years ago.

Helping people face what is true. Helping them hold together under pressure. Helping them decide that either we all climb the mountain, or none of us do.

Clarity

Imagine having apples in your kitchen.

Before you can eat one, you pay someone to collect the apples, drive them across town, load them onto a ship, send them to a factory, slice them, place them inside a very professional-looking box, and mail them back to your house.

Three days later, you open the package and say, “Hello there, apple. Have we met?”

It’s the same apple that was sitting six feet from you on Tuesday. It has simply returned with international shipping fees and a light coating of administrative despair.

The obvious question is:

Why would you do it in that order?

Why not take a knife, cut the apple yourself, and eat it?

Aliko Dangote became a billionaire by repeatedly asking a version of a question you’ve probably asked your spouse, your best friend, or the coworker who created a seven-step approval process for ordering printer paper:

Why are we doing it in this order?

For decades, Nigeria produced crude oil while depending heavily on imported gasoline, diesel, and other refined petroleum products. The raw material left the country. Much of the valuable processing happened elsewhere. The finished product then returned carrying additional costs.

Nigeria had the oil.

Foreign refineries had much of the processing capacity.

Nigeria bought back the finished product.

Hello there, oil. Have we met?

Dangote looked at that sequence and saw where money, expertise, control, and bargaining power were leaving the country. He invested roughly $20 billion in a refinery designed to process about 650,000 barrels of crude oil per day inside Nigeria. The facility is now operating and has strengthened Nigeria’s ability to refine fuel domestically, though prices and supply still depend on crude availability, exchange rates, distribution, policy, and global markets.

He built the missing middle.

That’s the idea I can’t stop thinking about.

We’re in love with invention. We celebrate the person who creates a category, discovers a technology, or announces an idea so new that everyone briefly loses the ability to use lowercase letters.

Every generation will have a few people who genuinely create something nobody has seen before.

But much of the working world doesn’t need a never-before-seen, generation-changing idea that immediately gets someone placed on the cover of TIME with their arms folded.

Sometimes it needs someone to look at what’s already there and ask:

What’s here because it works?

What’s here because it has always been here?

And what would happen if we changed the order?

Dangote’s genius wasn’t simply that he built something.

He saw a process that made little sense for the people living inside it, then rearranged who controlled the most valuable part.

Coherence

We tend to think of innovation as addition.

A new product. A new platform. A new department. A new strategic priority that arrives with seventeen slides, a logo, and the strong possibility that everybody’s calendar is about to get worse.

But some of the most consequential innovation is rearrangement.

Keep the elements.

Change the order.

Remove an unnecessary handoff.

Move the work closer to the person it’s supposed to benefit.

Make the path between contribution and outcome easier to see.

charity: water offers another example.

The organization didn’t invent charitable giving. It didn’t invent water. It looked at one of the questions sitting underneath nearly every donation:

Where did my money actually go?

How do I know the $20 I intended for someone who needs clean water didn’t disappear into executive expenses, administrative fog, or the coffers of a warlord?

That uncertainty weakens trust.

So charity: water redesigned the financial architecture around the donation. The organization promises that 100 percent of public donations will fund clean-water projects, while private donors, corporate partners, investment earnings, and donated goods support operating expenses. It maintains separate purposes, fundraising goals, and accounts for the two sides of the model.

The organization didn’t create a new human impulse.

People already wanted to give.

It changed the sequence surrounding that impulse so donors could draw a clearer line between the money they contributed and the outcome they hoped to produce.

That’s sequencing.

Every process contains inherited decisions. Some are intelligent. Some were reasonable when they were made. Some were temporary solutions that received permanent employment.

And some exist because a person who retired in 2014 once said, “Let’s do it this way for now,” and nobody has felt emotionally prepared to revisit the matter.

Inside organizations, the same pattern appears everywhere.

A university requires a student to contact four offices because each office owns one fragment of the answer.

A hospital asks a patient to repeat the same information to three people because the systems don’t speak to one another.

A leadership team requests frontline input after the strategy has already been approved.

A company trains employees to use AI before deciding which work the technology should improve.

Every individual step may appear reasonable.

The sequence is absurd.

Yet organizations often respond to a broken sequence by demanding better execution.

Work faster.

Communicate more.

Hold people accountable.

Add a dashboard.

But speed can’t rescue a process whose order is wrong. It only helps the organization complete the wrong sequence more efficiently.

This matters even more now.

AI makes execution cheaper and faster. It can draft the email, produce the deck, summarize the meeting, build the prototype, and create a 40-page implementation plan before anyone has finished asking whether the implementation plan should exist.

Execution is becoming less scarce.

Sequencing judgment is becoming more valuable.

The advantage may not belong to the person with the most ideas or the person who can produce the most work. It may belong to the person who can look at a familiar process and see that step seven should happen before step two.

Dangote didn’t invent petroleum.

He saw that Nigeria began with crude oil, ended with fuel, and paid an enormous economic penalty because too much of the valuable middle happened somewhere else.

He didn’t change the destination.

He changed who controlled the journey.

That distinction matters.

The refinery isn’t a magic wand. Domestic refining doesn’t automatically guarantee cheap fuel at every pump. Nigeria has continued to import some petroleum products and, at times, crude oil itself. Prices still respond to supply, exchange rates, policy, logistics, and global disruptions.

But the strategic logic remains powerful.

When a country produces the raw material but imports the finished product, the missing processing capacity isn’t a small operational inconvenience.

It’s where much of the money lives.

Dangote built himself into that gap.

Action

Some broken sequences are obvious.

Others hide inside language that sounds perfectly reasonable.

A leader hires someone, cuts a program, redirects resources, or rejects a proposal. When asked to explain the decision, they say:

“I believed it was the best decision.”

That sounds more sophisticated than “because I said so,” but it often performs the same function.

A credible decision has a sequence:

Evidence → Interpretation → Tradeoffs → Decision → Explanation

When a leader jumps from evidence directly to authority, the people affected can’t see how the decision was reached. They’re asked to accept the conclusion without being shown the reasoning that produced it.

Sometimes the organization doesn’t need a different decision.

It needs the missing steps restored.

That’s why leaders should ask a different question before creating another initiative, policy, product, department, or strategic priority:

Is something missing, or are the existing pieces simply in the wrong order?

Pick one outcome your organization claims to care about: student success, patient access, customer retention, employee development, donor trust, AI adoption, or faster decisions.

Then trace the actual sequence.

Where does the work begin?

Who touches it?

Where does it wait?

Where is information repeated?

Where does authority sit?

When does the person closest to the work get consulted?

Which step creates the outcome?

Which steps exist because they’re necessary?

Which ones exist because they’ve always been there?

The question I’d keep in front of the room is this:

Where are we exporting the work and importing the result?

An organization can do this without crossing a national border.

It exports judgment upward and imports decisions downward.

It exports employee development to an annual training event and imports disappointment when behavior doesn’t change.

It exports customer understanding to a survey vendor and imports a presentation explaining what the employees who speak with customers every day already knew.

It exports strategy to an executive retreat and imports a polished plan the operating system can’t use.

This is the kind of work I help leadership teams do when the organization has all the right pieces but still can’t consistently produce the outcome it wants.

We examine the sequence.

Where the work changes hands.

Where judgment disappears.

Where authority arrives too late.

Where an inherited step has become an expensive ritual.

Sometimes the answer isn’t another initiative.

It’s changing when the conversation happens, where the decision sits, who receives the information first, or how close authority is placed to the work.

That intervention can look boring.

No grand invention. No category creation. No magazine cover.

Just a process that finally makes sense for the people it was supposedly built to serve.

Boring can be worth billions.

Signal Boost

charity: water’s 100% Model is worth studying as an example of process redesign.

The organization didn’t answer donor skepticism only by explaining itself more persuasively. It changed the financial architecture so the promise became easier to verify.

That gives us a larger strategic principle:

When people don’t trust the promise, don’t only explain the promise more clearly. Redesign the process so the promise becomes easier to see.

For paid subscribers

The essay above makes one argument:

Some organizations don’t need a new strategy, a new tool, or another initiative.

They need the pieces they already have to happen in a better order.

The problem is that a broken sequence rarely announces itself as a broken sequence.

It shows up as delay.

Rework.

Mistrust.

Escalation.

A process that requires seven people to produce an answer one person already knew on Tuesday.

This week’s paid tool is The Sequence Reset, a 45-minute working session for identifying the exact point where a process stops making sense and redesigning it without automatically adding another person, platform, committee, or meeting.

By the end, you’ll have:

• One real process mapped as it actually happens

• The precise point where the sequence breaks

• One handoff to remove, move, or redesign

• One decision to place closer to the work

• A new sequence to test for 30 days

• A simple set of measures for determining whether the reset worked

Free readers get the argument.

Paid subscribers get the intervention.

Read more

Clarity

“Daddy, why can’t the government just print enough money for everyone?”

I started answering like a responsible adult.

“Well, because of the economy.”

Then I realized I had to explain the economy.

So I said something like, “The economy is everything that happens when people buy and sell goods and services.”

Then I heard myself say “goods and services” out loud and immediately realized I now had to explain the difference between goods and services.

So I tried again.

“Well, when a country produces things, that affects GDP, which stands for gross domestic product, and that means the total value of…”

At some point, I looked at this child, this tiny auditor from the Department of Humbling Questions, and said the only thing left to say.

“Who wants ice cream?”

You ever try to explain something to a child and realize halfway through that you don’t actually understand it as well as you thought?

You know the words.

You know which words usually sit next to each other.

Economy. Inflation. GDP. Money supply. Productivity. Interest rates. Very serious man on CNBC with a pocket square.

But then someone asks you to explain the thing plainly, and suddenly the whole concept starts leaking out of your hands.

And that matters.

Because if you don’t understand something deeply enough to explain it, how can you trust that you’re using it correctly, responsibly, or to its full potential?

When I finally did more research so I could answer my son’s question, something unexpected happened. I didn’t just get better at explaining the economy to a child. I got better at understanding the business side of my own consulting practice. Pricing made more sense. Value creation made more sense. Markets made more sense. The simple question forced me underneath the borrowed language.

That same story is being lived out in real time by leaders in every industry as their organizations grapple with whether, where, and how to implement AI.

They know the words.

Intelligence. Learning. Reasoning. Neural networks. Prediction. Automation. Transformation.

They’ve heard the language enough times to recognize it.

But recognition is not understanding.

And when leaders don’t understand what AI actually is, they either inflate it into magic or reduce it into a toy.

Neither response is useful.

There’s a simple reason this keeps happening:

Leaders are introduced to AI through the wrong metaphor.

They hear words like intelligence, learning, reasoning, and neural networks, and suddenly it starts to sound like there’s a tiny overcaffeinated consultant living inside the machine, reading the internet, connecting dots, and waiting politely for your prompt like it has a Patagonia vest, three unread McKinsey PDFs, and a strong opinion about “unlocking efficiency.”

That is not what is happening.

AI is not magic.

It is a pattern engine.

At the simplest level, generative AI learns from massive amounts of data, identifies relationships inside that data, and predicts what is most likely to come next. For language models, that means predicting the next word, or more accurately, the next token.

A token is a small piece of text. Sometimes it is a full word. Sometimes it is part of a word.

When you ask AI a question, it does not pull an answer from a shelf the way a person might pull a book from a library. It takes your prompt, reads the context, and begins generating an answer one piece at a time.

This word. Then the next. Then the next.

Each piece changes the context for the piece after it.

That is why you can put the same prompt into an AI tool multiple times and get slightly different answers. The model is not retrieving one fixed response from a folder labeled “Tadé asked about AI again.” It is generating a response through probabilities. There may be several reasonable next words, sentences, examples, or structures that fit the prompt. Sometimes it chooses one path. Sometimes it chooses another.

Same question.

Different possible continuation.

That variation is not a glitch. It is part of how the system works.

And this is where the phrase “prediction machine” is both true and incomplete.

Yes, AI is predicting what comes next. But prediction at this scale is not small. To predict well, the model has to learn patterns in grammar, facts, tone, structure, argument, code, analogy, genre, and reasoning. It has to learn that certain ideas tend to travel together. It has to learn how questions usually connect to answers, how explanations unfold, how stories move, how code functions, how arguments are built, and how people usually describe the world.

A quick resource for anyone who wants to go underneath the metaphor:

Stanford has a two-hour lecture explaining how large language models like ChatGPT and Claude are built, from transformers and training data to fine-tuning and the systems that turn a model into a usable AI product.

You do not need to become an engineer to lead around AI. But if you want a deeper understanding of what is happening beneath the interface, these are two hours worth investing.

Watch the full Stanford lecture here.

So yes, AI is a prediction machine.

But that phrase can make it sound like a very expensive autocomplete wearing a tiny graduation robe.

It is better to say this:

AI is a pattern engine trained through prediction.

It learns by trying to predict what comes next. When it gets the prediction wrong during training, the system adjusts. Over time, across enormous amounts of data, those adjustments create a model that can generate useful language, summarize information, write code, classify patterns, translate text, draft messages, and help people think through ideas faster.

That is the power.

It is also the problem.

Because AI predicts what fits.

It does not automatically know what is true.

A sentence can fit the pattern and still be wrong. An answer can sound polished and still be shallow. A recommendation can feel confident and still miss the context. A summary can look clean and still distort what mattered.

This is the part leaders need to understand.

AI is very good at fluency. It can make unclear thinking sound organized. It can make weak strategy sound plausible. It can make a half-formed idea look like a finished plan. It can take something that was spiritually held together with tape, vibes, and one exhausted program manager named Lisa, and turn it into a document with headings.

That does not mean the thinking got better.

It means the output got smoother.

Coherence

This matters because AI is entering organizations through speed.

It’s being introduced as a way to write faster, summarize faster, draft faster, analyze faster, respond faster, build faster, automate faster, and produce more with less friction.

That sounds good because speed is easy to defend.

Who wants to be the person in the room arguing against faster?

That person sounds like they still prints MapQuest directions and describes email as “the computer mail.” Nobody wants to be that person.

But speed is not the same as improvement.

A team can use AI to produce three times as many reports and still make worse decisions. A manager can use AI to write cleaner performance feedback and still avoid the conversation that actually matters. A department can generate more content, more summaries, more slide decks, more meeting notes, more strategy language, and more polished internal documents while the real work gets blurrier.

AI can make the work look more complete than it is.

That’s the danger.

It can make a weak idea easier to circulate. It can make an untested assumption sound mature. It can make a vague strategy feel finished because now it has structure, headers, and a conclusion that says something like “this approach positions us for sustainable success.”

Which, respectfully, may be true.

It may also be a fog machine in a blazer.

The issue is not that AI is bad. The issue is that AI increases the distance between fluency and judgment.

Before AI, a weak draft often looked weak. The gaps were visible. The clumsy parts announced themselves. The writer had to wrestle with the idea long enough to notice where the thinking broke.

Now the machine can smooth the broken parts before anyone has decided whether the idea deserves to live.

That changes the leadership problem.

The question is no longer only:

Can AI help us produce this?

The better question is:

Should this be produced?

And if it should be produced, what human judgment still has to protect the work?

That is where many AI conversations become too small. They stay trapped in tool selection, adoption rates, prompt libraries, policy language, and vendor promises. Those things matter, but they are not the center.

The center is judgment.

Who decides what good looks like?

Who checks whether the output is true?

Who knows enough about the work to see what the model missed?

Who has the authority to stop the workflow when the output is fluent but wrong?

Who can say, “This is faster, but it’s not better”?

Because that’s the sentence every organization will need sooner than it thinks.

This is why AI does not remove the need for judgment. It increases the cost of not having it.

Once a tool can produce more language, more summaries, more drafts, more plans, more slides, and more analysis at speed, the real question becomes harder to avoid:

Who is deciding what matters?

Who is checking whether the answer is true?

Who is asking whether the work got better?

Who is noticing whether we are creating return or simply producing more noise with a nicer font?

This is where the human work begins.

AI can generate possibilities, but humans still have to decide which possibilities are worth pursuing. AI can summarize the pattern, but humans still have to ask whether the pattern matters. AI can accelerate the work, but humans still have to decide whether the work should be accelerated.

That last part matters more than we admit.

Because some work should not be accelerated. Some work should be questioned. Some work should be redesigned. Some work should be stopped entirely, placed gently in a folder called “we tried,” and allowed to rest with its ancestors.

AI can help you move faster.

It cannot tell you whether faster is wise.

Action

Here is how this plays out inside a real organization.

A leadership team buys an AI platform. The announcement is strong. The rollout is professional. There is a training session with a title like “AI for Productivity,” which everyone attends while answering email because the irony has unionized.

A few months later, usage is up.

People are drafting faster. Meetings are being summarized. Internal updates look cleaner. Someone in operations is using AI to turn messy notes into polished project briefs. Marketing is producing more copy. Managers are generating first drafts of staff communications. The dashboard says adoption is increasing.

The room feels relieved.

Finally, movement.

But when you look closer, the question changes.

Are decisions getting better?

Are workflows improving?

Is quality increasing?

Are people saving meaningful time, or moving time into review, correction, and cleanup?

Are teams using AI to sharpen judgment, or to avoid the friction that judgment requires?

That is the work I help leadership teams do when AI activity starts outpacing AI measurement. The goal is not to shame experimentation. Experimentation matters. The goal is to separate leverage from theater before the organization starts scaling what it has not learned how to evaluate.

Because “people are using it” is not an AI strategy.

It is a usage report.

The harder work is building a measurement discipline around the tool. That means identifying where AI is touching the work, what human judgment still has to verify, what quality standards matter, where time is actually being saved, what new risks are being created, and which use cases should scale, stop, or be redesigned.

The practical move this week is simple.

Pick one AI use case in your organization, not the whole strategy.

One workflow.

One repeated task.

One place where people are already using the tool.

Then ask four questions:

What work is AI touching here?

What human judgment does this still require?

Did the work get better, or did it only get faster?

What would we measure in 30 days to know whether this created real return?

Those questions will tell you more than another broad conversation about AI readiness.

They will also reveal whether your organization is treating AI as leverage or theater.

And if the room cannot answer them, that is not failure.

That is information.

It means the organization may be moving faster than its measurement system, faster than its decision rights, and faster than its shared understanding of what “better” actually means.

That is where the work begins.

Signal Boost

A line to sit with:

AI predicts what fits. Humans decide what matters.

Use that as a diagnostic this week.

Where is AI helping judgment?

Where is it replacing judgment?

Where is it accelerating work that should probably be questioned first?

Where is it producing more output without producing more return?

That answer will tell you whether AI is becoming leverage or theater.

For paid subscribers

The essay above explains the basic idea: AI is not magic. It is a pattern engine trained through prediction.

The paid section gives you the operating tool.

This week’s tool is The AI Judgment Test, a practical diagnostic for leaders, teams, and operators who want to know whether AI is improving the work or just increasing output.

Free readers get the essay. Paid subscribers get the diagnostic.

Upgrade to get The AI Judgment Test.

Read more

Clarity

I remember being 18 and truly believing I was insanely mature. After all, I could drive a car. By myself. With no adults in the car.

I had an income too. (From my job at the mall, but still, a job is a job).

Then the 25-year-old version of me looked back at 18-year-old me like, wow, what a baby. But the 25-year-old version was different. This was maturity. I had a college degree. I paid bills. (Well, just my phone bill, but still. Give a brotha some credit.)

Then the 30-year-old version of me looked back at both of them and laughed. Then the married version of me in my mid-30s looked back at all of them like toddlers fighting over who was taller. And so forth.

Every version of us is tempted to believe the current version is the mature one.

Which raises an annoying question:

How do you actually know when maturation is happening?

If this were a philosophy newsletter, I might stop there and stare out a window for 900 words. But since we’re talking about leadership, clear thinking, decisions, and systems, I’ll rephrase it.

How do you tell whether an organization is actually maturing, instead of merely updating the dates on documents?

I think one answer is judgment.

But not judgment as a personal compliment. Not the kind of thing we say when someone has good instincts, makes smart decisions, or seems to see around corners. That matters, of course. I’d rather have a leader with good judgment than one treating the organization like a casino with a strategic plan.

But organizational maturity is not measured only by whether the person at the top has judgment. It is measured by whether judgment can travel.

Can judgment move from the executive suite to the front desk? Can it move from the strategic plan to the program manager? Can it move from the values page to the employee who has thirty seconds to make a call the policy didn’t anticipate?

That’s where maturity shows up.

Not in how impressive the language is.

In whether people can use it.

Coherence

At some point, the CEO has to decide whether to chase the shiny new technology that’ll make the board feel like everyone owns matching Patagonia vests and a Gartner login, or invest in the dependable thing that won’t make the news but will actually make the organization work better.

That’s judgment.

But if that’s the only place judgment exists, the organization is fragile. Everything has to climb. Every exception escalates. Every weird situation gets kicked upstairs. Every frontline person learns to follow the process even when the process is clearly failing the moment in front of them.

That’s not discipline. That’s a very expensive permission slip factory.

A mature organization doesn’t reserve judgment for senior leaders. It teaches and empowers judgment at every level. The front desk admin is not simply checking people in and checking boxes. They are noticing when the person in front of them needs a different kind of help than the script anticipated. The program manager is not only moving tasks from one column to another like a tiny bureaucratic Roomba. They are paying attention to where the plan and the actual work have started to separate. The new employee is not blindly following instructions. They understand enough of the mission to know when something deserves attention.

That is where mission, vision, and purpose (MVP) either become real or stay decorative.

Because MVP is not supposed to be a paragraph on the website. It is supposed to become judgment.

The CEO should understand what the mission requires from the executive seat. The front desk admin should understand what the mission requires from the first point of contact. The program manager should understand what the mission requires when the project plan starts lying.

Different seats. Different decisions. Same clarity.

That’s the part many organizations miss. They communicate mission, vision, and purpose as language, but they don’t translate it into judgment. So people can repeat the words, but they can’t use them. They can name the values, but they can’t apply them under pressure. They can recite the strategy, but they still have to ask permission for every call that requires discernment.

This is where organizations shrink without noticing. The org chart says people are trusted. The workflow says, “Please escalate anything that requires a pulse.”

A healthy organization builds clarity into the system. It doesn’t depend on one brilliant person at the top seeing everything, deciding everything, and rescuing everyone from ambiguity with a calendar that looks like it was assembled by raccoons.

Judgment has to be taught. People need to know what good judgment means here. What gets protected. What gets prioritized. What tradeoffs are acceptable. What standards don’t bend just because the room got tense.

And judgment has to be empowered. People need enough trust to use it before everything climbs the org chart and dies somewhere between “circling back” and “let’s revisit next quarter.”

Action

Here’s how this plays out inside a real organization.

A leadership team spends months clarifying the strategic plan. The language gets better. The retreat feels productive. The document looks polished enough to make everyone briefly believe the organization has been reborn through font choice.

Then the plan meets the actual work.

A student-facing employee has to decide whether to follow the process exactly or solve the problem in front of them. A manager has to decide whether to protect the timeline or name that the timeline was built on fantasy. A director has to decide whether to escalate an issue or handle it locally. A vice president has to decide whether the stated priority is real enough to disappoint someone important.

This is where you find out whether the organization has matured. Because if everyone is still waiting for permission, the plan has not become judgment. It has become decoration with bullet points.

The test is not whether people can quote the mission. The test is whether they can interpret it from their seat.

When I work with leadership teams, this is one of the places I pay attention to closely. Strategy often sounds aligned at the top while judgment breaks down in the middle and at the edges. People are not always confused because they lack commitment. Sometimes they are confused because the organization has not clarified what good judgment looks like when the situation is messy, political, urgent, or human.

That work requires more than a better statement. It requires translation.

What does our mission require when two priorities collide? What does our vision require when speed and quality are fighting? What does our purpose require from the person closest to the work? Which decisions should climb, and which ones should be trusted closer to the ground?

That’s the work.

Not making the language prettier.

Making the judgment transferable.

Because if clarity only lives at the top, the system doesn’t really have clarity. It has a few people making calls for everyone else.

Signal Boost

Peter Drucker is often credited with the line, “Culture eats strategy for breakfast.”

It’s a great line. I know this because I can’t take five steps on LinkedIn without seeing it somewhere. Usually on top of a stock photo of a chess piece, a mountain, you get the vibe.

I digress.

But I think the comparison is apples to oranges.

Strategy is not the stated direction. Strategy is the unique value an organization can create that customers, students, patients, clients, or communities can’t get the same way from anyone else. It is the thing the organization is supposed to obsess over until you can see it everywhere: in what gets offered, how the service feels, what sales promises, what marketing repeats, how budgets get built, what the operating model protects, and what people decide when nobody senior is in the room.

That last part matters more now than it used to.

In a world of AI-everything, almost anyone can execute almost anything at lightning speed. Which means good judgment is more valuable than ever, because it keeps you from producing trash at scale when you were trying to scale effectiveness.

But judgment is also something else. It is a litmus test of organizational maturity. It’s one way you can look back and say, yes, we got the projects done, but we’re not actually maturing. Or, yes, that project failed, but look at the maturation it forced. Look at what we learned to see. Look at the judgment we built that will help us win more consistently over time.

Culture is different. Culture is the environment that decides what people are actually allowed to protect, challenge, notice, ignore, name, and decide.

So I don’t think culture eats strategy.

I think culture converts strategy.

Sometimes it converts strategy into discipline. Sometimes it converts strategy into theater. Sometimes it converts strategy into a beautiful PDF that everyone praised, filed, and then worked around until the next planning cycle.

A strategy can say, “Our unique value is high-touch student support.” But if the culture rewards speed over care, people will process students instead of support them.

A strategy can say, “Our unique value is deep expertise.” But if the culture rewards volume over judgment, people will start producing more and thinking less.

A strategy can say, “Our unique value is trust.” But if the culture punishes local decision-making, every meaningful call will climb the org chart like it’s trying to summit Everest in business casual.

That’s why mission, vision, and purpose (MVP) can’t stay trapped in language. They have to become judgment.

Because the real test isn’t whether people can repeat the strategy. The real test is what the culture turns the strategy into once people have to use it under pressure.

Does it become better judgment?

Or does it become prettier language wrapped around the same old defaults?

For paid subscribers

The essay above names the problem: an organization is not mature just because the language is clear at the top. It is mature when judgment can travel.

This week’s paid tool is The Maturity Test: Can Judgment Travel Here?

Use it to examine whether your mission, vision, purpose, or strategy has become usable judgment across the organization, or whether people are still waiting for permission when the work gets messy.

Free readers get the insight. Paid readers get the tool behind the insight.

Read more

I don’t think people hate meetings as much as they say they do.

I think they hate being invited into a room that doesn’t fulfill the expectation the invite created.

It’s like lunch.

I may not be hungry. I may be completely fine not eating right now.

But if you tell me you’ll be at my office in five minutes and you’re bringing lunch, then show up with a fortune cookie and some Tic Tacs, now I’m disappointed.

Not because I was starving.

Because you created an expectation and didn’t deliver what you led me to expect.

That’s what a lot of meetings do.

The invite says one thing.
The room does another.

And people don’t just respond to the meeting.

They respond to the gap.

We usually categorize meetings by content. This meeting’s about sales. This meeting’s about workflow. This meeting’s about strategy. This meeting’s about planning.

That sounds reasonable because content matters. People should know what the topic is. But the topic doesn’t tell people what kind of meeting they’re walking into.

And that’s where a lot of meeting culture starts to break.

The real problem is a lack of differentiation between content and kind. It’s possible, and sadly probable, to have meetings with drastically different content and identical formats. A sales meeting, a workflow meeting, a strategy meeting, and a team meeting can all end up being the same meeting with different words dropped into the agenda.

Different content.

Same room.
Same rhythm.
Same vague opening.
Same wandering middle.
Same “great conversation” ending where nothing actually changed.

That’s not a meeting strategy.

That’s a reusable container with no discernment.

Different kinds of meetings have different purposes, in the same way different tools serve different functions. When you need to drive a nail into wood, you use a hammer. When you need to turn a screw into place, you use a screwdriver. When you need to stir soup, you use a ladle. When you need to serve food, you use a spoon.

The problem is that many leaders keep using one meeting tool for very different jobs.

They use a hammer when they need a screwdriver.
They use a hammer when they need a ladle.
They use a hammer when they need a serving spoon.

And yes, sometimes they actually need a hammer.

That’s part of what makes the problem so sticky.

Because when the hammer works once, leaders take that as proof that the hammer’s the right tool. But it may only be proof that one job required a hammer. It’s not proof that every job does.

That’s what happens when leaders use the same meeting kind for different meeting purposes. The format works just enough to make them trust it. Then it fails often enough to make everyone else dread it.

Clarity

The number one thing stopping leaders from having better meetings is that we categorize meetings by content when we should also categorize them by kind.

A bad meeting’s often a misalignment problem before it’s a time management problem.

The meeting type isn’t clear. The structure doesn’t match the type. The goal doesn’t match the structure.

So people walk in carrying different assumptions about what the hour’s supposed to produce.

Some people think they’re there to decide. Some people think they’re there to connect. Some people think they’re there to process tension. Some people think they’re there to give updates and avoid eye contact until the calendar releases them back into the wild.

This is why meetings get weird.

People talk past each other, but not always because they’re bad communicators. Sometimes they’re participating in different meetings at the same time.

One person’s trying to close the loop. Another person’s trying to build trust. Another person’s trying to name the thing nobody wants to say. Another person’s wondering why this couldn’t have been an email, a voice note, or a carrier pigeon with a PDF attached.

The leader thinks, “That went fine.”

The team thinks, “I’ve got no idea what just happened, but somehow I’ve got three more tasks and one more feeling.”

That’s not a meeting problem.

That’s a clarity problem wearing a calendar invite.

There are three basic meeting types leaders need to distinguish: transactional, relational, and adaptive.

Transactional meetings are about getting work done. Standups, sales updates, project check-ins, planning sessions, execution reviews, decision meetings. These meetings need visible structure. They need shared documents, owners, deadlines, decisions, and a record of what changed because the meeting happened.

A transactional meeting should make work clearer than it was before the meeting started. If people leave with more confusion than they entered with, something’s gone wrong. The point isn’t to admire the work from several angles. The point’s to move it.

Relational meetings are different. They’re about trust, context, and connection. Onboarding conversations. Team retreats. Leadership offsites. Culture sessions. Moments when people need to know one another as more than job titles with laptops.

Relational meetings shouldn’t be treated like calendar filler. Strong relationships are infrastructure. But relational meetings still need structure. They need prompts, listening, reflection, and enough room for people to choose how much of themselves they want to bring forward.

That last part matters.

You can ask someone to map their career pivots. You can ask them to name one piece of work-related advice they return to often. You can ask them to tell a story of resilience.

But the point isn’t forced vulnerability.

Nobody needs a surprise group therapy ambush at 9:03 on a Tuesday.

The point’s laddered openness. People need enough structure to participate without feeling exposed, cornered, or tricked into becoming content for someone else’s culture deck.

Adaptive meetings are different again. These are for complex, sensitive, uncertain, or high-stakes issues where the right process or outcome isn’t obvious at the beginning.

An adaptive meeting says, “Something unexpected happened. What do we do now?”

Basically every meeting in April 2020 was an adaptive meeting. Nobody was just running the usual agenda. People were trying to make decisions with incomplete information, emotional pressure, real consequences, and no clean playbook.

That’s adaptive work.

Strategy sessions live here. So do innovation sprints, career conversations, cross-functional conflict, organizational responses to major societal events, and leadership conversations where everyone can feel the tension but nobody’s found a clean way to name it yet.

Adaptive meetings require agility, safety, and real facilitation. They need structure, but not rigidity. They need truth, but not recklessness. They need enough pace to keep the work moving and enough patience to let the real issue surface.

If transactional meetings are about execution, and relational meetings are about trust, adaptive meetings are about coherent movement through uncertainty.

That distinction matters because each meeting type asks people to show up differently.

When leaders don’t make the meeting type explicit, people bring their own assumptions.

And assumptions are where meeting culture starts charging interest.

Coherence

Imagine I show up ready for a transactional meeting.

I’ve reviewed the numbers. I know the blockers. I’ve got my updates ready. I’m prepared to make decisions, assign owners, and leave with next steps.

Then the leader opens with:

“Before we get into anything, I’d love for everyone to tell the story of their leadership journey.”

That might be a beautiful prompt.

In the right meeting.

But in this meeting, it lands like someone moved the exit door.

Because I came prepared to produce outcomes.

So now I’m sitting there trying to adjust my face into “deeply present and emotionally available” while my brain’s still holding six unresolved operational issues and a customer escalation that’s chewing through the drywall.

Nothing’s technically wrong with the prompt.

That’s the annoying part.

It may be thoughtful. It may be useful. It may even be exactly what the team needs at some point.

But because the meeting was framed one way and run another way, the prompt gets interpreted through irritation.

“They don’t respect my time.”
“This meeting’s pointless.”
“Leadership’s out of touch.”
“We’re avoiding the real work.”

Now flip it.

I show up expecting a relational meeting. Maybe the invite says team connection, trust-building, or leadership alignment. I arrive prepared to listen, reflect, and engage with people in a more human way.

Then the leader opens with:

“What did you get done?”
“What’s the plan?”
“What’s the timeline?”
“Why isn’t this finished?”

Those may be fair questions.

In the right meeting.

But now they land as evaluation, not alignment. The leader may think they’re being focused. The team may experience them as cold, callous, or bait-and-switchy.

Accountability isn’t the issue.

The issue is that the meeting was framed as connection and then run as inspection.

That mismatch creates a different story.

“It’s not safe to be honest here.”
“They only care about productivity.”
“This trust stuff is fake.”
“I should keep my guard up.”

Same meeting room. Same people. Same leader. Completely different interpretation because the meeting type, structure, and goal were fighting each other.

That’s why meeting clarity isn’t political correctness, corporate theater, or LinkedIn thought leadership filler.

It’s what keeps people from spending the whole meeting privately translating the leader’s intent.

A meeting isn’t just time on a calendar. It’s a temporary operating system. It tells people how to think, what to notice, what to protect, what to risk, and what kind of contribution is appropriate.

If the operating system’s unclear, people start guessing.

And when people guess, they usually guess in the direction of self-protection.

They speak less honestly. They prepare for the wrong thing. They hold back the useful objection. They wait to see what the leader really wants. They leave with private interpretations instead of shared commitments.

That’s how meetings become expensive without looking expensive.

The cost doesn’t always show up as a bad meeting score. It shows up as delayed decisions, duplicated work, passive resistance, vague ownership, relational residue, and the exhausted sentence every organization knows too well:

“Didn’t we already talk about this?”

Yes.

You talked about it.

But the meeting wasn’t designed to produce what you needed from it.

Action

The fix starts before the meeting begins.

Leaders need to name the meeting type explicitly. Not just the topic. The type.

You don’t need to make this dramatic. You don’t need to turn the calendar invite into a novella. You just need to tell people what kind of work the room’s being asked to do.

“We’re here to decide and assign next steps.”

That’s transactional.

“We’re here to build trust and strengthen alignment.”

That’s relational.

“We’re here to work through something sensitive and real where the next move isn’t obvious yet.”

That’s adaptive.

Saying this at the start of the meeting helps. But the stronger move is to signal it in advance.

“For next week’s meeting, come prepared to make a decision on the vendor.”

“For next week’s meeting, come prepared to discuss what’s been getting in the way of trust across the team.”

“For next week’s meeting, come prepared to sit with a messy question. We may not leave with a final answer, but we do need to leave with a clearer read on the tension.”

You can even put these expectations directly into the calendar invite so people have a ready-made place to find them before they walk into the room.

That one sentence changes how people prepare. They know what to bring, what to expect, what to protect, and what kind of honesty the meeting’s asking from them.

Then the structure needs to match the type.

A transactional meeting needs an agenda, a shared document, a decision log, owners, deadlines, and clear next steps. If nobody can say what changed because the meeting happened, the meeting probably didn’t earn its time.

A relational meeting needs prompts, reflection, listening, and enough psychological room for people to participate without being dragged onto an emotional stage they didn’t audition for.

An adaptive meeting needs safety, structure, slower pacing, careful facilitation, the courage to name tension, and enough discipline to avoid rushing toward fake closure because silence made someone uncomfortable.

Using the wrong tools makes the meeting fight itself.

It’s like asking someone to open up while staring at them like they just backed into your parked car.

Your words are requesting trust.

Your face has filed a police report.

Leaders don’t need to make every meeting deeper, warmer, faster, or more efficient. They need to make each meeting more honest about what it is.

That’s the work.

When a leadership team says, “Our meetings aren’t working,” I usually don’t start with the agenda. I start with the operating logic underneath the meeting.

What kind of meeting is this supposed to be? What does it need to produce? What are people assuming when they walk in? What structure would actually support that outcome? Where’s the current format creating friction, avoidance, or false agreement?

This is the work I help leadership teams do when meetings have become a place where alignment is performed but not produced.

Because better meetings are rarely about better calendar hygiene.

They’re about sharper leadership judgment.

A good meeting earns the time it takes. It tells people why they’re there, gives them the right tools for the work, and produces something the organization can actually use.

That something may be a decision. It may be trust. It may be a clearer way through uncertainty.

But if the leader hasn’t named which one, the team will guess.

And guesswork is where meeting culture starts to rot.

Signal Boost

Priya Parker writes in The Art of Gathering that “specificity is a crucial ingredient.”

That applies directly here.

The more generic the meeting, the more people rely on habit. The more specific the purpose, the more people can bring the right version of themselves into the room.

A meeting called “Team Sync” asks people to guess.

A meeting framed as “Decide the Q3 implementation owner and unblock the timeline” gives people a job.

A meeting framed as “Rebuild trust after the handoff failure” gives people a different job.

A meeting framed as “Make sense of the strategy tension before we commit resources” gives people another one.

Meeting types aren’t about having more efficient meetings for the sake of efficiency.

They’re about respecting people’s time by giving the highest probability that when stressed-out, busy people come together, they’re doing something that adds value to their day, to their projects, and to them.

Clarity

We have a bunch of stuff from my son’s soccer team that keeps moving around the house. At first, it was in the kitchen. Then, when we were cleaning the kitchen, it got moved to the living room. When we were cleaning because guests were coming over, it got moved to the office. Then, when we were cleaning the office, it got moved to the guest room.

Spoiler: it’s back in the kitchen as of this afternoon.

That’s funny until you realize a lot of organizations do the exact same thing with findings, concerns, risks, and uncomfortable information. They don’t ignore it. They acknowledge it, discuss it, capture it in notes, and say something mature like, “This is really important,” or “We need to come back to this,” or “Let’s revisit it next quarter.”

Which is corporate for: please put this uncomfortable thing somewhere it can’t hurt our calendar.

That’s Insight Parking.

Insight parking happens when an organization recognizes something true, but moves it somewhere else instead of deciding what to do about it. It’s not denial. Denial says, “That’s not true.” Insight parking says, “That’s true, which is exactly why we need to discuss it later, with the right stakeholders, after alignment, once priorities settle, ideally during a month that does not exist.”

The problem is that insight parking feels responsible. A concern was named, a finding was validated, a note was captured, and someone probably said “great discussion,” which is how you know the meeting had already started losing oxygen.

But nothing changed.

No owner, no deadline, no tradeoff, no decision, and no consequence for letting the finding sit there until everyone gets used to stepping around it.

That’s where organizations start to lose trust.

Coherence

A lot of organizations don’t suffer because no one knows what’s wrong. They suffer because people know exactly what’s wrong, but the organization has developed elegant ways to avoid converting that knowledge into action.

The customer feedback already said the process is confusing. The employee survey already said people don’t trust senior leadership. The equity audit already said certain groups are carrying more friction than others. The strategy retreat already surfaced the fact that everyone says yes in the room and then optimizes for their own department after the meeting. The frontline staff already warned that the new system would break at the exact point where it later broke.

None of this is hidden.

That’s what makes it so expensive.

Organizations love to talk about blind spots, but many of the most damaging problems are not blind spots. They are named spots. Documented spots. Already-discussed spots. We-have-a-slide-on-that spots.

The issue is not awareness. It’s conversion

Can the organization convert what it knows into a decision?

That is where things get hard, because decisions create consequences. Once you decide, someone owns something. Once someone owns something, priorities shift. Once priorities shift, something else loses status, time, money, or protection.

So the organization parks the insight. It moves the soccer bag from the kitchen to the living room, then from the living room to the office, then from the office to the guest room, until one day everyone is standing in the kitchen again asking, “Wait, why is this still here?”

Because moving something is not the same as resolving it.

This is one of the reasons good people stop bringing their best thinking into rooms. It’s not always because they’re disengaged. Sometimes they’re doing the math. They watched the organization recognize the right thing, watched leadership nod, watched the concern get validated, and watched nothing happen.

After a while, people learn the real rule: insight is welcome as long as it doesn’t require a decision.

That rule destroys trust. Not all at once. It happens slowly. A little less honesty, a little less urgency, a little less effort to name what everyone can already see. People still attend the meetings. They still smile at the right moments. They still type “Thanks!” in the chat like emotionally exhausted citizens of a tiny failing republic.

But their best thinking is gone because the organization taught them that insight has no path.

Action

The first move is not to ask, “What do we know?”

Most organizations already know more than they’re willing to act on.

The better question is:

What have we already acknowledged that still has no owner, deadline, or decision attached to it?

That’s the parked work.

In a leadership team, I’d start by making the parked things visible. Not as a complaint session or a blame exercise, but as a decision inventory. What findings keep resurfacing, what concerns keep getting validated but not resolved, what risks are being discussed in careful language because nobody wants to name the tradeoff, and what keeps getting moved to the next meeting, next quarter, next planning cycle, next version of the strategy, next leader, next miracle?

Then I’d separate the parked items into three categories: things we are ready to decide, things we are not ready to decide but need a clear owner and timeline, and things we are pretending to revisit but actually need to kill.

That third category matters.

A lot of organizational drag comes from issues that are kept alive because nobody wants to say no. So they get honored into permanent ambiguity. They become sacred clutter.

And sacred clutter is still clutter.

The leadership work is not to act on every insight. That would be chaos with a calendar invite. The leadership work is to make the path clear: this gets decided, this gets owned, this gets tested, this gets killed, and this gets revisited on a real date with a real person responsible for bringing a recommendation.

This is the work I help leadership teams do when good thinking keeps getting acknowledged but not converted into decisions.

Because the expensive part is not always the problem itself.

The expensive part is the loop.

The same finding returns, the same concern resurfaces, the same friction gets renamed, and the same people get asked for candor, then watch their candor get professionally stored somewhere harmless.

Eventually, the organization doesn’t just lose time.

It loses the people who were still willing to tell the truth.

So here’s the question:

What has your organization already named but not decided?

Find that.

That’s probably where the real work starts.

Signal Boost

A useful decision rule:

If the finding is real enough to acknowledge, it’s real enough to assign.

That does not mean every finding needs immediate action. But it does mean every meaningful finding deserves a clear next step: decide, own, test, schedule, or kill.

Anything else is probably just insight parking with better lighting.

Clarity

You ever see a dumpster fire right at the moment it gets lit?

It’s memorable.

One second you’re enjoying your day.

Next moment, everything smells like trash.

And now you can’t finish the sandwich you’re holding because it too smells like trash.

That’s what it felt like watching an executive commission an expensive report, receive the findings, and then convene a committee to figure out what to do with them.

Not a small report.

A serious one. (Serious is my euphemism for very expensive).

Lots of interviews, data, etc.

Which sounds normal until you sit with it for five seconds.

Shouldn’t the order be reversed?

Shouldn’t the sequence be:

We need to make a decision.

We don’t yet have enough evidence to make it well.

Let’s get the data we need.

Then, when the evidence comes back, we decide.

That would make sense. That would be clean. That would be adult behavior, which is always nice to encounter in the wild, like seeing a raccoon use a napkin.

But that’s not what happened.

They got someone to collect a mountain of data. Then they started asking who else they should bring in to help them understand what to do with the data.

At one point, they even considered hiring a second consultant to interpret the first consultant’s findings.

That’s when I realized the report was never the problem.

The data was not unclear.

The decision was.

And when the decision is unclear, more data does not create courage.

It creates fog with citations.

Coherence

Organizations love data because data feels responsible.

It gives the room something to point at. It creates the feeling of seriousness. It lets people say, “We’re being evidence-based,” while still avoiding the more dangerous sentence:

“We have to choose.”

A report can be useful. A good one can sharpen the room. It can surface patterns people were too close to see. It can reveal gaps, contradictions, risks, and opportunities. I’m not against reports. I’ve written them. I’ve used them. I have deep respect for a clean finding that walks into a room and destroys a bad assumption.

But data is supposed to serve a decision.

When the decision has not been named, data starts serving the avoidance.

That is when the organization enters the loop.

Commission the report.

Review the findings.

Form the committee.

Ask for more context.

Request another analysis.

Add stakeholders.

Schedule the follow-up.

Decide the issue is complex.

Commission something else.

Boom. Everyone has been responsible, nobody has been accountable, and the original decision is still sitting there like a diner at table 12 wondering if the server forgot they exist.

This is not a data problem.

It is a courage problem wearing a blazer.

The deeper issue is that many leadership teams don’t ask the first question early enough:

What decision is this evidence supposed to help us make?

Without that question, the report has no job. It becomes an object of interpretation instead of a tool for commitment.

A good report should create pressure. It should make the next responsible decision harder to ignore. But if the room has no decision frame, the same report becomes a professionally formatted hiding place.

This is where data insight and people insight have to work together.

Data insight tells you what the evidence reveals.

People insight tells you what the evidence threatens.

Sometimes the evidence threatens a strategy. Sometimes it threatens someone’s credibility. Sometimes it threatens a budget, a timeline, a public commitment, or a version of the story leadership has been telling itself for too long.

That is why organizations sometimes gather more information when what they really need is to face the cost of acting on what they already know.

The issue is not that they don’t have enough data.

The issue is that the data has started asking for a decision they don’t want to make.

Action

Inside a real organization, this usually shows up in one of three ways.

The first is the report with no decision owner. Everyone agrees the findings matter, but nobody can say whose decision the findings are supposed to change.

The second is the report with too many owners. Every stakeholder gets a voice, which sounds inclusive, until ownership becomes so diluted that the finding loses force.

The third is the report with a hidden decision. The organization says it wants to “understand the data,” but what it really wants is permission to delay, soften, bury, or outsource the decision the data has made harder to avoid.

This is the point where I’d stop the room and ask:

What decision is actually on the table?

What are we pretending is still unclear?

What would become different if we accepted the finding?

Who has the authority to act?

Who pays if we do nothing?

Those questions change the work.

They move the room from interpretation to consequence.

This is the work I help leadership teams do when an expensive report, strategic finding, data pattern, or stakeholder concern has created more discussion than movement.

We don’t start by asking, “What does the data say?”

We start by asking:

What decision is this evidence trying to force?

Then we map what the finding threatens, who carries the cost, what kind of resistance is showing up, and what action now has to become visible.

Because a leadership team does not need another beautiful PDF if the real problem is decision avoidance.

It needs structure.

It needs pressure.

It needs someone to help the room stop admiring the evidence and start deciding what the evidence requires.

Before your next report, audit, survey, assessment, or consultant engagement, ask one question:

What decision will this help us make?

If the answer is vague, don’t buy the report yet.

Name the decision first.

Otherwise, you may end up with 80 pages, six meetings, a committee, a second consultant, three new acronyms, and the emotional texture of a conference room sandwich.

All of it documented.

None of it decided.

Signal Boost

Tool: The Decision-First Data Brief

Before commissioning new research, write one page with four sentences:

1. The decision we need to make is:

2. The evidence we do not yet have is:

3. The tradeoff we are trying to understand is:

4. When the evidence comes back, the decision owner will be:

If those four sentences can’t be written clearly, the organization may not need more data yet.

It may need a clearer decision.

I’ve spent my whole career in higher ed and academia, so my writing can be a bit stoic.

Training will do that to you.

Be objective. Be careful. Do not sound too excited.

But today? I’m excited.

Straight-up excited y’all!

I finally scratched an itch I’ve been thinking about for some time.

Every company with an AI strategy can usually tell you its carbon footprint.

Almost none can tell you its AI footprint.

That gap is not an accident.

Carbon accounting had decades to mature. Treaties. Reporting standards. Consultants. Dashboards. Entire compliance functions built around CO2.

AI usage has none of that yet. There is no common unit, no shared language, and no clean way to compare one organization’s usage against another’s.

So I built one.

AEU: AI Environmental Unit.

One AEU represents the environmental cost of one million standard AI inference calls on the average U.S. grid. It accounts for model size, query complexity, training overhead, grid carbon intensity, and water usage.

You can access the AEU calculator here: https://aeu.clarityovernoise.co

I didn’t build this because I think AI is bad.

I use AI. I study AI. I build with AI. I’m not outside the building shaking a clipboard at the future like the neighborhood association president of innovation.

The issue is simpler than that.

AI is becoming infrastructure. And infrastructure has costs.

Right now, most organizations are scaling AI adoption without a serious way to measure one of those costs: the environmental load created by their usage.

That will not hold.

Every board asking, “What is our AI strategy?” is one budget cycle away from a regulator, investor, employee, customer, or journalist asking, “What is your AI cost to the planet?”

Most organizations do not have an answer. They have vendor language, productivity language, “responsible AI” language, and maybe a task force, because when institutions get scared, they immediately form a group of twelve people with a shared folder and no actual authority.

But they do not have a number.

And when there is no number, there is no shared argument.

Measurement Changes Behavior

Carbon accounting did not fix climate change.

But it changed the room.

It gave companies, regulators, investors, employees, and customers a shared way to see impact and compare behavior. Once emissions could be measured, they could be reported. Once they could be reported, they could be challenged. Once they could be challenged, they became part of strategy, regulation, capital, and reputation.

That is what measurement does when it works. It turns concern into structure. It gives an issue a way into rooms where moral urgency alone gets converted into “we’ll revisit this next quarter,” which is corporate for “please walk this idea into the woods.”

AI needs that same shift.

Right now, two companies can both say they are “using AI responsibly” while doing completely different things. One company might use lightweight tools for targeted workflows: summaries, routing, documentation, internal support. Another might run frontier models across every function, generate huge volumes of prompts, train custom systems, process massive datasets, and do it through infrastructure with serious energy and water demands.

On paper, both can look innovative.

Environmentally, they are not the same organization.

But if the balance sheet has no column for that difference, the difference disappears.

That is how invisible costs win. Not because leaders are evil, but because systems reward what they can see.

Why AEU Matters

AEU is a first attempt to give AI usage an environmental measurement language.

It is not the final word. It should not be. Any first framework that claims perfection should be asked to leave the meeting and drink water.

The point is to create something usable enough to argue with, improve, and build on. A useful metric does not need to answer every question. It needs to force better ones.

Which AI use cases are worth their environmental cost? Which tasks require larger models, and which ones do not? Which teams are using high-intensity tools for low-stakes work? Which vendors can report meaningful energy and water data? Which parts of an AI strategy create real benefit, and which parts are just novelty with a procurement code?

That last one matters.

A lot of AI adoption right now is less strategy and more corporate raccoon behavior: shiny tool, grab it, run into the ceiling tiles, make noise.

Measurement helps leaders separate serious use from decorative use.

That distinction is going to matter because real regulation does not start with outrage. It starts with measurement.

Outrage can create pressure, but regulation needs units, thresholds, categories, reporting norms, and enforceable claims. Without measurement, the conversation stays trapped in mood: AI is dangerous, AI is inevitable, innovation matters, sustainability matters, and please enjoy this panel discussion where seven people agree to define nothing.

Measurement gives the argument a spine.

The Leadership Issue

This is not only an environmental issue.

It is a leadership issue.

AI strategy is not just about tools. It is about judgment under pressure. It is about what an organization chooses to scale, what it chooses to ignore, and what it is willing to measure before someone else measures it for them.

That is where this connects to the work I do with leadership teams.

When a strategy has momentum but the hidden costs have not been named, the task is not to slow everything down. The task is to build enough clarity that speed does not become drift.

For AI, that means helping leaders name the real decision. Are we adopting AI everywhere, or only where it improves the work? Are we reducing cost, increasing capacity, improving quality, protecting human judgment, or building a governance model that can survive contact with reality?

Those are different strategies.

They require different measures.

AEU is one measure for one part of the problem. It will not solve AI governance by itself. No metric deserves that much power. We have all suffered enough under dashboards built by people who think numbers are a personality.

But a good metric can change the conversation. It can make invisible costs visible, help leaders compare choices, create pressure before the crisis point, and give boards, employees, customers, investors, and regulators a shared way to ask better questions.

That is the point.

Not to shame companies for using AI.

To make AI adoption responsible…through measurement.

To normalize this question:

Does your organization know its AI footprint?

Or would it be guessing?

The AEU framework is live now: https://aeu.clarityovernoise.co

I think this conversation is about to get much bigger, and I’m excited to help shape it.

If your organization is thinking seriously about AI strategy, governance, environmental impact, or the hidden costs of adoption, this is the work I help leadership teams do: make the invisible costs visible before they become strategic risk.

You can learn more about working with me here.

A team can hit its goals and still hurt the organization.

That sounds dramatic, but it happens every day. A department improves its numbers by making everyone else’s work harder. A leader protects their budget by starving a cross-functional priority. A team optimizes for speed and creates downstream chaos for the people responsible for quality, implementation, or customer trust.

On paper, the subsystem is winning.

In reality, the system is paying for it.

Donella Meadows calls this suboptimization: when the goals of one subsystem dominate at the expense of the goals of the total system.

It’s one of those ideas that makes you want to stare out the window for a minute like you’re in a documentary about municipal plumbing.

Because once you see it, you can’t unsee it.

Clarity

Suboptimization happens when one part of the organization becomes too successful at pursuing its own goals. That’s the part people miss. The problem isn’t always incompetence. Sometimes the problem is competence pointed in the wrong direction.

A sales team closes deals the delivery team can’t fulfill. A finance team cuts costs in ways that weaken morale, capacity, or trust. A product team ships faster by pushing complexity onto customer support. A senior leader demands alignment, but rewards individual departments for defending their own territory.

Everyone can explain their behavior. Everyone can justify their metrics. Everyone can say, with a clean conscience and possibly a very expensive slide deck, “We were doing what we were asked to do.”

That’s what makes suboptimization so dangerous. It usually looks reasonable up close.

But systems don’t care about your department’s intentions. They respond to structure, incentives, feedback loops, and goals. If the goals are fragmented, the behavior will be fragmented. If every team is rewarded for protecting its own numbers, don’t be surprised when collaboration starts to look like a group project where everyone is secretly praying someone else does the bibliography.

The deeper issue is this:

Misaligned goals create rational behavior that produces irrational outcomes.

That’s why leaders can’t only ask, “Is this team performing?”

They also have to ask, “What is this team’s performance costing the larger system?”

Because sometimes the win is real.

And so is the damage.

Coherence

I’ve seen this play out in organizations where every department was full of smart, committed people.

That’s the annoying part.

It would be easier if dysfunction came with villains. A person in a cape. A VP dramatically laughing while deleting the onboarding plan. Usually, it’s not that. Usually, it’s a room full of people doing their best inside a system that rewards the wrong things.

One team is told to move faster. Another team is told to reduce risk. Another team is told to protect quality. Another team is told to keep costs down. Each goal makes sense by itself. Together, they create tension nobody has named.

So the organization starts producing symptoms.

Projects stall after approval. Meetings multiply because decisions don’t hold. People agree in the room, then re-litigate the decision in smaller rooms. Leaders say “we need better communication,” when what they really have is goal conflict dressed up as a communication problem.

That last one matters.

Because if you misdiagnose suboptimization as a communication issue, you’ll keep asking people to collaborate harder inside a structure that teaches them not to. You’ll add updates. You’ll add dashboards. You’ll add another meeting, because apparently we as a species have learned nothing and will continue putting calendar invites on top of structural problems like frosting on a traffic cone.

But more communication won’t fix misaligned goals.

More central control won’t always fix it either.

Meadows warns about this too. Too much local control can fragment the system. Too much central control can suffocate it. Healthy systems need enough hierarchy to coordinate, enough self-organization to adapt, and enough resilience to survive pressure without collapsing into blame, compliance, or chaos.

This is where suboptimization becomes more than a systems problem.

It becomes a risk problem.

Once leaders see that one part of the organization is winning at the expense of the whole, the next question is not only, “How do we fix it?”

It’s, “What does admitting this now cost?”

A department may know its behavior is hurting the whole. But admitting that may mean:

Logic Risk: “Our way of measuring success is flawed.”

Strategic Risk: “The initiative we funded and announced is producing the wrong behavior.”

Social Risk: “The leaders who designed this system now have to admit they helped create the problem.”

That’s why suboptimization survives.

Not because people can’t see it.

Because seeing it clearly creates a bill someone has to pay.

Sometimes the risk is logical: the insight breaks the operating model people have trusted for years. Sometimes it’s strategic: the organization has already invested money, reputation, and political capital in the current direction. Sometimes it’s social: naming the problem makes specific leaders look wrong in front of peers.

That deepens the leadership challenge.

The work is not only to identify the misalignment.

The work is to reduce the cost of telling the truth about it.

That means leaders have to get underneath the surface questions. Who is rewarded for what? Where do team goals compete with system goals? Where are people solving local problems by creating larger ones? Where does “accountability” mean protecting the department instead of serving the mission? Where has the organization confused control with coherence?

These questions show up in budgets, meetings, performance reviews, reporting lines, project delays, handoffs, and the strange emotional weather of an organization.

You can feel when a system is over-controlled.

You can feel when a system is under-coordinated.

And you can definitely feel when everyone is technically aligned but spiritually preparing for a knife fight in Q3 planning.

Action

This is the kind of work I do with leadership teams when strategy looks aligned on paper, but execution keeps stalling in practice.

The issue is rarely that leaders don’t care. The issue is usually that the system is asking different parts of the organization to win in ways that make the whole organization lose.

A useful starting point is to map the misalignment tax.

Pick one strategic priority that has been slow, frustrating, or politically expensive. Then ask:

1. Which teams have to cooperate for this to work?

2. What does each team believe it is being rewarded to protect?

3. Where do those goals reinforce each other?

4. Where do they compete?

5. What behavior makes perfect sense locally, but creates damage systemwide?

6. What risk would leaders face if they admitted the misalignment out loud?

That sixth question matters.

Because the barrier is not always awareness.

Sometimes people already know. They know the metric is distorting behavior. They know the initiative is creating the wrong incentives. They know the handoff is broken. They know the current structure rewards local wins and systemwide loss.

But knowing is not the same as being able to say it in the room.

That’s where facilitation matters.

When I work with leadership teams, part of the work is helping them name the misalignment without turning the room into a courtroom. The goal is not to make someone the villain. The goal is to help the system tell the truth without everybody immediately grabbing a shield, a spreadsheet, and a lawyer they found on LinkedIn.

That’s where the leverage is.

Not in asking people to try harder.

In redesigning the conditions that shape how they act.

If this dynamic feels familiar inside your organization, this is the work I help leadership teams do: identify the misalignment tax, surface the competing goals, reduce the risk of naming what’s true, and build a more coherent path from strategy to execution.

Because a system can’t sustain itself when every part is winning alone.

Eventually, the whole pays the bill.

Signal Boost

Donella Meadows, Thinking in Systems:

“When a subsystem’s goals dominate at the expense of the total system’s goals, the resulting behavior is called suboptimization.”

That sentence is worth carrying into your next strategy meeting.

Especially if everyone’s numbers look good, but the organization still feels harder to move than it should.

—Forwarded this email? Subscribe here for more

When One Team Wins and the System Loses

Tadé Ayeni

Jun 23

READ IN APP

A team can hit its goals and still hurt the organization.

That sounds dramatic, but it happens every day. A department improves its numbers by making everyone else’s work harder. A leader protects their budget by starving a cross-functional priority. A team optimizes for speed and creates downstream chaos for the people responsible for quality, implementation, or customer trust.

On paper, the subsystem is winning.

In reality, the system is paying for it.

Donella Meadows calls this suboptimization: when the goals of one subsystem dominate at the expense of the goals of the total system.

It’s one of those ideas that makes you want to stare out the window for a minute like you’re in a documentary about municipal plumbing.

Because once you see it, you can’t unsee it.

Clarity

Suboptimization happens when one part of the organization becomes too successful at pursuing its own goals. That’s the part people miss. The problem isn’t always incompetence. Sometimes the problem is competence pointed in the wrong direction.

A sales team closes deals the delivery team can’t fulfill. A finance team cuts costs in ways that weaken morale, capacity, or trust. A product team ships faster by pushing complexity onto customer support. A senior leader demands alignment, but rewards individual departments for defending their own territory.

Everyone can explain their behavior. Everyone can justify their metrics. Everyone can say, with a clean conscience and possibly a very expensive slide deck, “We were doing what we were asked to do.”

That’s what makes suboptimization so dangerous. It usually looks reasonable up close.

But systems don’t care about your department’s intentions. They respond to structure, incentives, feedback loops, and goals. If the goals are fragmented, the behavior will be fragmented. If every team is rewarded for protecting its own numbers, don’t be surprised when collaboration starts to look like a group project where everyone is secretly praying someone else does the bibliography.

The deeper issue is this:

Misaligned goals create rational behavior that produces irrational outcomes.

That’s why leaders can’t only ask, “Is this team performing?”

They also have to ask, “What is this team’s performance costing the larger system?”

Because sometimes the win is real.

And so is the damage.

Coherence

I’ve seen this play out in organizations where every department was full of smart, committed people.

That’s the annoying part.

It would be easier if dysfunction came with villains. A person in a cape. A VP dramatically laughing while deleting the onboarding plan. Usually, it’s not that. Usually, it’s a room full of people doing their best inside a system that rewards the wrong things.

One team is told to move faster. Another team is told to reduce risk. Another team is told to protect quality. Another team is told to keep costs down. Each goal makes sense by itself. Together, they create tension nobody has named.

So the organization starts producing symptoms.

Projects stall after approval. Meetings multiply because decisions don’t hold. People agree in the room, then re-litigate the decision in smaller rooms. Leaders say “we need better communication,” when what they really have is goal conflict dressed up as a communication problem.

That last one matters.

Because if you misdiagnose suboptimization as a communication issue, you’ll keep asking people to collaborate harder inside a structure that teaches them not to. You’ll add updates. You’ll add dashboards. You’ll add another meeting, because apparently we as a species have learned nothing and will continue putting calendar invites on top of structural problems like frosting on a traffic cone.

But more communication won’t fix misaligned goals.

More central control won’t always fix it either.

Meadows warns about this too. Too much local control can fragment the system. Too much central control can suffocate it. Healthy systems need enough hierarchy to coordinate, enough self-organization to adapt, and enough resilience to survive pressure without collapsing into blame, compliance, or chaos.

This is where suboptimization becomes more than a systems problem.

It becomes a risk problem.

Once leaders see that one part of the organization is winning at the expense of the whole, the next question is not only, “How do we fix it?”

It’s, “What does admitting this now cost?”

A department may know its behavior is hurting the whole. But admitting that may mean:

Logic Risk: “Our way of measuring success is flawed.”

Strategic Risk: “The initiative we funded and announced is producing the wrong behavior.”

Social Risk: “The leaders who designed this system now have to admit they helped create the problem.”

That’s why suboptimization survives.

Not because people can’t see it.

Because seeing it clearly creates a bill someone has to pay.

Sometimes the risk is logical: the insight breaks the operating model people have trusted for years. Sometimes it’s strategic: the organization has already invested money, reputation, and political capital in the current direction. Sometimes it’s social: naming the problem makes specific leaders look wrong in front of peers.

That deepens the leadership challenge.

The work is not only to identify the misalignment.

The work is to reduce the cost of telling the truth about it.

That means leaders have to get underneath the surface questions. Who is rewarded for what? Where do team goals compete with system goals? Where are people solving local problems by creating larger ones? Where does “accountability” mean protecting the department instead of serving the mission? Where has the organization confused control with coherence?

These questions show up in budgets, meetings, performance reviews, reporting lines, project delays, handoffs, and the strange emotional weather of an organization.

You can feel when a system is over-controlled.

You can feel when a system is under-coordinated.

And you can definitely feel when everyone is technically aligned but spiritually preparing for a knife fight in Q3 planning.

Action

This is the kind of work I do with leadership teams when strategy looks aligned on paper, but execution keeps stalling in practice.

The issue is rarely that leaders don’t care. The issue is usually that the system is asking different parts of the organization to win in ways that make the whole organization lose.

A useful starting point is to map the misalignment tax.

Pick one strategic priority that has been slow, frustrating, or politically expensive. Then ask:

1. Which teams have to cooperate for this to work?

2. What does each team believe it is being rewarded to protect?

3. Where do those goals reinforce each other?

4. Where do they compete?

5. What behavior makes perfect sense locally, but creates damage systemwide?

6. What risk would leaders face if they admitted the misalignment out loud?

That sixth question matters.

Because the barrier is not always awareness.

Sometimes people already know. They know the metric is distorting behavior. They know the initiative is creating the wrong incentives. They know the handoff is broken. They know the current structure rewards local wins and systemwide loss.

But knowing is not the same as being able to say it in the room.

That’s where facilitation matters.

When I work with leadership teams, part of the work is helping them name the misalignment without turning the room into a courtroom. The goal is not to make someone the villain. The goal is to help the system tell the truth without everybody immediately grabbing a shield, a spreadsheet, and a lawyer they found on LinkedIn.

That’s where the leverage is.

Not in asking people to try harder.

In redesigning the conditions that shape how they act.

If this dynamic feels familiar inside your organization, this is the work I help leadership teams do: identify the misalignment tax, surface the competing goals, reduce the risk of naming what’s true, and build a more coherent path from strategy to execution.

Because a system can’t sustain itself when every part is winning alone.

Eventually, the whole pays the bill.

Signal Boost

Donella Meadows, Thinking in Systems:

“When a subsystem’s goals dominate at the expense of the total system’s goals, the resulting behavior is called suboptimization.”

That sentence is worth carrying into your next strategy meeting.

Especially if everyone’s numbers look good, but the organization still feels harder to move than it should.

This Substack is reader-supported. To receive new posts and support my work, consider becoming a free or paid subscriber.

Dr. Tadé Ayeni is a scholar, author, and executive advisor. He developed the Insight Creates Risk framework, which helps senior leaders diagnose why good insight stalls and navigate it into practical outcomes. He serves as a director and assistant professor at Washington State University and consults with senior leaders across North America, Europe, and Africa. He is currently completing doctoral research on how leaders navigate complex, AI-mediated organizations. His book, Clarity of Thought, was published in 2025.

Work With Me · clarityovernoise.co

Clarity

The traditional way to pick a World Cup national team is simple.

Find the best goalkeeper.
Find the best right back.
Find the best center backs.
Find the best midfielders.
Find the best forwards.
Put them on the plane.
Let the nation argue about the one winger who got robbed.

It is clean.

It is democratic.

It makes sense to fans, pundits, group chats, radio hosts, former players, taxi drivers, and every man in a pub wearing a retro shirt who has said “I’m just saying” before committing a minor tactical hate crime.

The logic is obvious.

Best players.
Best team.

But Thomas Tuchel appears to be making a different bet with England.

He has left out big names. Obvious names. Players with talent, reputation, and highlight reels that look like they were edited by someone who drinks espresso from a thimble and says “cinema” too much.

Cole Palmer.
Phil Foden.
Trent Alexander-Arnold.
Harry Maguire.

Depending on your preferred group chat argument, one of those omissions probably made you sit up straighter.

The easy read is that Tuchel is being ruthless.

The better read is that he is challenging the operating logic.

He said it plainly:

“I think from day one we were very clear that we are trying to select and build the best possible team, which is not necessarily to select and collect the 26 most talented players.”

Instead of asking, “Who are the best 26 English players?”

He’s choosing to ask, “Which 26 players give me the best chance of building the team I actually want to play?”

That is a different question.

And different questions create different risks.

I’ve named this logic risk.

It happens when a better read on the situation threatens the assumptions people have been using to make decisions.

The old logic says:

Pick the best players.

The new logic says:

Pick the players who make the system work.

That sounds simple until a star gets left at home.

Coherence

Every organization has a selection logic.

Sometimes it is official.

Most of the time, it lives in the walls.

This is how we pick leaders.
This is how we promote people.
This is how we decide who gets invited into the room.
This is how we define “ready.”
This is how we know who matters.

In football, the selection logic is usually performance plus reputation plus position.

If a player is brilliant for his club, he should go.

If he has been one of the best players in the country, he should go.

If he plays a position where England needs quality, he should go.

That logic feels fair because it rewards visible performance.

But international football does not give you much time.

You do not get nine months to install culture.
You do not get a long preseason.
You do not get to solve chemistry with a task force, three listening sessions, and a 47-slide deck titled “Creating Alignment Through Shared Excellence,” which should be illegal under several international treaties.

You get a few weeks.

So the question changes.

Can this player accept the role?
Can he perform without the whole structure being built around him?
Can he make the team more coherent, or does his talent require the team to bend?
Can he sit, enter, adjust, press, cover, and still be fully bought in?
Can the group absorb him without becoming less itself?

That is process over performance.

Performance tells you who has been excellent.

Process tells you what that excellence requires from everyone else.

We saw a version of this with Argentina in 2022.

Yes, they had Messi.

That part helps.

I have heard that having one of the greatest footballers alive is occasionally useful, especially when the alternative is asking a committee to define “attacking identity” while everyone slowly loses the will to hydrate.

But Argentina did not win the World Cup by simply collecting the most gifted players and hoping genius would solve the wiring.

They built a team that allowed Messi to be Messi.

Runners around him.
Workers behind him.
Midfielders who could cover ground.
Defenders who could survive pressure.
Players who understood that their job was not to compete with the star, but to complete the system.

That was the logic shift.

The question was not, “Who has the most talent?”

The question was, “What structure gives the central idea its best chance to work?”

That is what makes Tuchel’s England selection interesting.

He may be wrong about the names.

But the question he is asking is the right kind of dangerous.

This is why Tuchel’s decision is not simply a selection debate.

It is a logic risk.

Logic risk appears when new insight contradicts the operating framework people have built everything on.

The old framework says: take the most talented players.

The new framework says: talent only matters if it survives the system you are trying to play.

That sounds obvious until a gifted player is left at home.

Then it becomes dangerous.

Because once you choose fit over fame, you are no longer protected by consensus.

If Tuchel picks the biggest names and England lose, he can hide inside the national ritual.

“We had the talent.”
“The players did not perform.”
“It came down to moments.”
“We go again.”

Everyone knows the script.

It comes laminated.

But if he leaves stars at home and England lose, the failure has a name tag.

It becomes his judgment.
His philosophy.
His risk.

That is the part leaders underestimate.

A familiar decision gives you cover.

An original decision gives you ownership.

And ownership is expensive.

This is where the England squad becomes bigger than football.

Organizations do this all the time.

They say they want collaboration, then promote the individual hero.

They say they want systems thinking, then reward the person who saves the broken process at midnight.

They say they want innovation, then staff the work with people who are safest on paper.

They say they want team performance, then keep selecting for individual shine.

And when someone challenges that logic, the room gets tense.

Because the question is no longer, “Who is best?”

The question becomes, “Best for what?”

That question exposes things.

It exposes whether your strategy is real or decorative.

It exposes whether your values can survive a hard choice.

It exposes whether the organization actually believes in the system it claims to be building.

Tuchel’s squad is risky because it forces that question into public.

If England win, the logic changes overnight.

Suddenly everyone will talk about balance, cohesion, specialist roles, emotional discipline, and the maturity to build a tournament team rather than a fantasy XI.

People will say it was obvious.

People love saying something was obvious after someone else paid the cost of proving it.

If England lose, the old logic will return with a machete.

Why leave out Palmer?
Why leave out Foden?
Why leave out Trent?
Why overthink it?

That is logic risk at full volume.

The new frame has to win before it is trusted.

The old frame gets trusted because it is old.

Action

The lesson for leaders is not “ignore talent.”

Please do not print that on a mug and begin ruining your succession planning process by Thursday.

The lesson is that every selection decision reveals the logic underneath the organization.

When a senior team chooses who leads the initiative, who enters the room, who gets promoted, who gets protected, or who becomes the face of the work, it is never only choosing a person.

It is choosing a philosophy.

Are we selecting for individual brilliance?
Role fit?
Trust?
Adaptability?
Political safety?
The ability to work inside the process without warping it?

This is exactly the work I do with leadership teams when strategy looks aligned on paper but execution keeps stalling.

The issue is often not the plan.

It is the selection logic behind the plan.

The wrong people may be in the wrong roles for reasons everyone can defend individually, but nobody has tested collectively.

One person has the strongest résumé.
One person has the most history.
One person has the loudest sponsor.
One person has been waiting their turn.
One person looks safest to the board.

Each decision makes sense on its own.

Then the team gets built.

And the process starts leaking.

The room has talent, but no rhythm.

The work has owners, but no coherence.

The strategy has names attached, but the system cannot move.

That is where I help leaders slow the room down enough to ask the question that usually gets skipped:

What are we actually selecting for?

Because if you do not name the logic, the default logic wins.

And the default logic is rarely neutral.

It protects what has already been rewarded.

It protects the people who already look legitimate.

It protects performance you can point to, even when the next challenge requires a different kind of fit.

Tuchel may be wrong.

That is the risk.

But the more interesting thing is that he has made the logic visible.

He has said, in effect:

I am not picking the safest argument.

I am picking the team I think can work.

That is the kind of decision that either gets mocked early or copied later.

Sometimes both.

Signal Boost

A useful question for any leader building a team:

Are you picking the best people individually, or the people who make the work most likely to succeed collectively?

Those are not always the same group.

And the difference between them is where a lot of strategy goes to die.

(P.S. I’m now invested in England doing well because I need this article to age well.)

Forwarded this email? Subscribe here for more

The Best Players Don’t Always Make the Best Team

Hear Me Out

Tadé Ayeni

Jun 17

READ IN APP

Clarity

The traditional way to pick a World Cup national team is simple.

Find the best goalkeeper.
Find the best right back.
Find the best center backs.
Find the best midfielders.
Find the best forwards.
Put them on the plane.
Let the nation argue about the one winger who got robbed.

It is clean.

It is democratic.

It makes sense to fans, pundits, group chats, radio hosts, former players, taxi drivers, and every man in a pub wearing a retro shirt who has said “I’m just saying” before committing a minor tactical hate crime.

The logic is obvious.

Best players.
Best team.

But Thomas Tuchel appears to be making a different bet with England.

He has left out big names. Obvious names. Players with talent, reputation, and highlight reels that look like they were edited by someone who drinks espresso from a thimble and says “cinema” too much.

Cole Palmer.
Phil Foden.
Trent Alexander-Arnold.
Harry Maguire.

Depending on your preferred group chat argument, one of those omissions probably made you sit up straighter.

The easy read is that Tuchel is being ruthless.

The better read is that he is challenging the operating logic.

He said it plainly:

“I think from day one we were very clear that we are trying to select and build the best possible team, which is not necessarily to select and collect the 26 most talented players.”

Instead of asking, “Who are the best 26 English players?”

He’s choosing to ask, “Which 26 players give me the best chance of building the team I actually want to play?”

That is a different question.

And different questions create different risks.

I’ve named this logic risk.

It happens when a better read on the situation threatens the assumptions people have been using to make decisions.

The old logic says:

Pick the best players.

The new logic says:

Pick the players who make the system work.

That sounds simple until a star gets left at home.

Coherence

Every organization has a selection logic.

Sometimes it is official.

Most of the time, it lives in the walls.

This is how we pick leaders.
This is how we promote people.
This is how we decide who gets invited into the room.
This is how we define “ready.”
This is how we know who matters.

In football, the selection logic is usually performance plus reputation plus position.

If a player is brilliant for his club, he should go.

If he has been one of the best players in the country, he should go.

If he plays a position where England needs quality, he should go.

That logic feels fair because it rewards visible performance.

But international football does not give you much time.

You do not get nine months to install culture.
You do not get a long preseason.
You do not get to solve chemistry with a task force, three listening sessions, and a 47-slide deck titled “Creating Alignment Through Shared Excellence,” which should be illegal under several international treaties.

You get a few weeks.

So the question changes.

Can this player accept the role?
Can he perform without the whole structure being built around him?
Can he make the team more coherent, or does his talent require the team to bend?
Can he sit, enter, adjust, press, cover, and still be fully bought in?
Can the group absorb him without becoming less itself?

That is process over performance.

Performance tells you who has been excellent.

Process tells you what that excellence requires from everyone else.

We saw a version of this with Argentina in 2022.

Yes, they had Messi.

That part helps.

I have heard that having one of the greatest footballers alive is occasionally useful, especially when the alternative is asking a committee to define “attacking identity” while everyone slowly loses the will to hydrate.

But Argentina did not win the World Cup by simply collecting the most gifted players and hoping genius would solve the wiring.

They built a team that allowed Messi to be Messi.

Runners around him.
Workers behind him.
Midfielders who could cover ground.
Defenders who could survive pressure.
Players who understood that their job was not to compete with the star, but to complete the system.

That was the logic shift.

The question was not, “Who has the most talent?”

The question was, “What structure gives the central idea its best chance to work?”

That is what makes Tuchel’s England selection interesting.

He may be wrong about the names.

But the question he is asking is the right kind of dangerous.

This is why Tuchel’s decision is not simply a selection debate.

It is a logic risk.

Logic risk appears when new insight contradicts the operating framework people have built everything on.

The old framework says: take the most talented players.

The new framework says: talent only matters if it survives the system you are trying to play.

That sounds obvious until a gifted player is left at home.

Then it becomes dangerous.

Because once you choose fit over fame, you are no longer protected by consensus.

If Tuchel picks the biggest names and England lose, he can hide inside the national ritual.

“We had the talent.”
“The players did not perform.”
“It came down to moments.”
“We go again.”

Everyone knows the script.

It comes laminated.

But if he leaves stars at home and England lose, the failure has a name tag.

It becomes his judgment.
His philosophy.
His risk.

That is the part leaders underestimate.

A familiar decision gives you cover.

An original decision gives you ownership.

And ownership is expensive.

This is where the England squad becomes bigger than football.

Organizations do this all the time.

They say they want collaboration, then promote the individual hero.

They say they want systems thinking, then reward the person who saves the broken process at midnight.

They say they want innovation, then staff the work with people who are safest on paper.

They say they want team performance, then keep selecting for individual shine.

And when someone challenges that logic, the room gets tense.

Because the question is no longer, “Who is best?”

The question becomes, “Best for what?”

That question exposes things.

It exposes whether your strategy is real or decorative.

It exposes whether your values can survive a hard choice.

It exposes whether the organization actually believes in the system it claims to be building.

Tuchel’s squad is risky because it forces that question into public.

If England win, the logic changes overnight.

Suddenly everyone will talk about balance, cohesion, specialist roles, emotional discipline, and the maturity to build a tournament team rather than a fantasy XI.

People will say it was obvious.

People love saying something was obvious after someone else paid the cost of proving it.

If England lose, the old logic will return with a machete.

Why leave out Palmer?
Why leave out Foden?
Why leave out Trent?
Why overthink it?

That is logic risk at full volume.

The new frame has to win before it is trusted.

The old frame gets trusted because it is old.

Action

The lesson for leaders is not “ignore talent.”

Please do not print that on a mug and begin ruining your succession planning process by Thursday.

The lesson is that every selection decision reveals the logic underneath the organization.

When a senior team chooses who leads the initiative, who enters the room, who gets promoted, who gets protected, or who becomes the face of the work, it is never only choosing a person.

It is choosing a philosophy.

Are we selecting for individual brilliance?
Role fit?
Trust?
Adaptability?
Political safety?
The ability to work inside the process without warping it?

This is exactly the work I do with leadership teams when strategy looks aligned on paper but execution keeps stalling.

The issue is often not the plan.

It is the selection logic behind the plan.

The wrong people may be in the wrong roles for reasons everyone can defend individually, but nobody has tested collectively.

One person has the strongest résumé.
One person has the most history.
One person has the loudest sponsor.
One person has been waiting their turn.
One person looks safest to the board.

Each decision makes sense on its own.

Then the team gets built.

And the process starts leaking.

The room has talent, but no rhythm.

The work has owners, but no coherence.

The strategy has names attached, but the system cannot move.

That is where I help leaders slow the room down enough to ask the question that usually gets skipped:

What are we actually selecting for?

Because if you do not name the logic, the default logic wins.

And the default logic is rarely neutral.

It protects what has already been rewarded.

It protects the people who already look legitimate.

It protects performance you can point to, even when the next challenge requires a different kind of fit.

Tuchel may be wrong.

That is the risk.

But the more interesting thing is that he has made the logic visible.

He has said, in effect:

I am not picking the safest argument.

I am picking the team I think can work.

That is the kind of decision that either gets mocked early or copied later.

Sometimes both.

Signal Boost

A useful question for any leader building a team:

Are you picking the best people individually, or the people who make the work most likely to succeed collectively?

Those are not always the same group.

And the difference between them is where a lot of strategy goes to die.

(P.S. I’m now invested in England doing well because I need this article to age well.)

Dr. Tadé Ayeni is a scholar, author, and executive advisor. He developed the Insight Creates Risk framework, which helps senior leaders diagnose why good insight stalls and navigate it into practical outcomes. He serves as a director and assistant professor at Washington State University and consults with senior leaders across North America, Europe, and Africa. He is currently completing doctoral research on how leaders navigate complex, AI-mediated organizations. His book, Clarity of Thought, was published in 2025.

Work With Me · clarityovernoise.co

For the first five years of my professional career, I believed a fairy tale.

For context, I spent years helping leaders at every level find better insight. When others said the data showed we should go left, I specialized in taking a different angle, finding the nuance, and revealing that we may actually need to go right.

I thought, like a lot of smart people think, that if I could get to the best insight, the rest would take care of itself.

Then something peculiar happened.

Once.

Then again.

Then again.

Until it stopped looking like an anomaly and started looking like the pattern.

Like clockwork, the same sequence repeated. Analyze the data. Flip it over. Analyze it again from a different perspective. Look at the quantitative and the qualitative. Find the one crucial insight that could move the project forward and help the organization get to the next level.

Then the insight would meet the pressure of the real world.

And instead of getting used, it would get talked about.

Or half-used.

Or changed until it became unrecognizable and useless.

Or ignored.

Or better yet, the very people who asked for it would get angry when you gave it to them because it didn’t support the narrative they’d already built.

The organization would keep limping along saying, “We need more data. We need more clarity. We need more stakeholder input. We need more time.”

All while the thing they said they needed was sitting right there.

You asked for it.

We delivered it.

You’re not using it.

Eventually, if you’re wired like me, you start asking a very reasonable question:

What sort of weird organizational Twilight Zone am I living in?

That frustration changed my work.

I stopped focusing only on how to find the best insight and started studying what it takes for the best insight to survive contact with real organizational pressure.

That’s where this essay begins.

Clarity

The problem is rarely the insight.

Smart organizations usually aren’t short on good thinking. They have data. They have talented people. They commission assessments. They run diagnostics. They hire consultants with tasteful slide decks and shoes that say, “I have airport status.”

The insight exists.

What they often lack is a way to move it.

That’s because good information doesn’t enter an organization as information.

It enters as implication.

If the insight is true, something else may have to change. A strategy may need to shift. A leader may need to walk back a public commitment. A team may need to admit that the metric they trusted was measuring the wrong thing. A department may need to stop defending a model that everyone has built language, budget, and identity around.

That’s why the gap between knowing something and acting on it usually isn’t a gap in intelligence.

It’s not simply a communication problem, though communication will always volunteer to take the blame because it has poor boundaries.

It’s not even, usually, a culture problem in the vague way people use that word when no one wants to say power, status, or “someone important made a decision that aged like yogurt in a hot car.”

It’s a cost problem.

Accepting good insight is expensive.

Not always financially expensive.

Usually socially expensive. Politically expensive. Identity-expensive.

Expensive in the way that requires someone to sit in a meeting and say, with their whole face:

“The thing I defended last quarter may not be the thing we need now.”

And until someone names what the insight costs, even the most accurate, well-researched, beautifully presented insight will find a way to stall.

This is what I call the Insight Creates Risk framework.

The simple version is this:

Data insight tells you what the evidence reveals.

People insight tells you what the evidence threatens.

You need both.

Without data insight, leaders drift into opinion, instinct, politics, and noise. Without people insight, leaders assume the facts should be enough.

They’re not.

Clear evidence can still die in the room if nobody understands what accepting it would cost.

Coherence

There was a program that should not have survived its first review.

A team had spent months building something genuinely new. The design was sound. The early results were promising. The work had shape. It had momentum. It had the faint smell of something that might actually matter.

Which, in organizational life, is usually the exact moment someone reaches for a spreadsheet and starts loading the crossbow.

By any honest measure, it was rare work. The kind of work that could become the flagship. The proof point. The thing people mention in rooms they weren’t invited to, which is, as we know, the highest form of institutional success.

Then they evaluated it.

They used the same criteria they’d always used.

The rubric built for the old model.

The model this program had been designed to move beyond.

The measuring stick from the house they’d supposedly moved out of but were still emotionally storing furniture in.

By those standards, the program underperformed.

The numbers said mediocre.

And since the numbers said mediocre, the room was ready to have an unnecessary moment of humility, admit the program had failed, throw it away, and move on to something “better.”

Which would’ve been very mature.

Also completely wrong.

Because by updated criteria, by the standards that matched what the program was actually trying to accomplish, it was exceptional.

It was succeeding at the thing everyone had claimed they wanted.

And the program nearly died anyway.

The issue wasn’t the work.

The issue was the cost of protecting it.

That cost had very little to do with budget. Very little to do with evidence. Almost everything to do with what accepting the new assessment would’ve required of the people in that room.

I’ve seen this pattern repeat enough times to stop treating it as strange.

A strong finding enters the room. It should change the conversation. Instead, it activates risk.

Above the surface, people ask clarifying questions.

Below the surface, the furniture is on fire.

In this case, three risks were working at the same time.

1. Logic Risk

Logic risk activates when new insight contradicts the operating framework a leader or organization has built everything on.

This is bigger than a single belief. It’s not only, “We thought the program would work and now we’re not sure.”

It’s about the system of assumptions that has justified the decisions, investments, structures, metrics, roles, meetings, reporting lines, and solemn little strategic phrases people have been repeating for years.

When insight threatens that system, accepting it means admitting the old logic may no longer hold.

And once coherence takes on a life of its own, it stops asking whether it’s still useful.

Even bad coherence can become self-protective. Especially bad coherence. Because once enough people have built careers around it, the logic becomes less like a tool and more like a family heirloom.

It may not work.

But everyone agrees it has to stay in the dining room.

In the program story, this was the first risk in the room. The team had built its assessment infrastructure around one model of what “good” looked like. The new program operated on different premises entirely.

Accepting the new criteria wasn’t a small rubric update.

It meant admitting the old rubric, and the decisions built around it, could no longer be fully trusted.

That’s an enormous ask.

People rarely volunteer to saw through the floor they’re standing on, even if the floor is made of laminated mission statements and vibes.

2. Strategic Risk

Strategic risk activates when acting on new insight means publicly unwinding a commitment someone has already made.

This is where organizations get tight in the shoulders.

Because organizations announce things. Leaders stake their reputations on directions. Budgets get approved. Initiatives get launched. Names get attached. Champions emerge.

Someone makes a kickoff speech with a sentence like:

“This is not just a program. This is who we are becoming.”

And now everyone has to go along with it, because nobody wants to be remembered as the person who made the one leader with promotion power look wrong in public.

Once that happens, new insight becomes dangerous.

Because if the insight says the direction is wrong, accepting it requires someone to absorb the cost of being wrong in front of the people who watched them commit.

That’s not a data problem.

That’s a public reversal problem.

In the program story, prominent leaders had already committed to how the program would be assessed. Changing the criteria mid-process wasn’t just a methodological adjustment.

It was a reversal.

Someone would have to say:

“The standard I put my name on was the wrong standard.”

Even in healthy organizations, that’s hard.

In unhealthy ones, it’s basically asking someone to walk barefoot across a conference table while the finance team takes notes.

So the organization reaches for softer language.

“We need more time.”

“We should revisit the assumptions.”

“We may not have enough evidence yet.”

“Let’s socialize this with a few more stakeholders.”

Which is often organizational code for:

Please give my ego a medically supervised exit ramp.

3. Social Risk

Social risk activates when accepting insight means absorbing a status cost in front of others.

This is the most interpersonal risk. It’s also the most underestimated, mostly because serious people like to pretend status isn’t real while arranging their entire lives around who gets copied on which email.

Organizations are social systems. Status, credibility, belonging, and reputation are real currencies. They shape what people say, what they avoid, what they defend, and what they pretend not to understand.

When insight threatens to make a leader look incompetent, uninformed, outdated, or complicit in something that didn’t work, the social cost of accepting it can outweigh the organizational benefit of acting on it.

In the program story, the leaders who built the original assessment criteria were still in the room.

Accepting the new framework publicly named their work as inadequate.

Even if no one said it directly, everyone would know.

And “everyone would know” is one of the most powerful governance mechanisms in organizational life.

No one has to accuse you.

The room just develops a weather system.

A pause. A glance. A careful sentence. Someone says “interesting” in a tone that means absolutely not.

That social exposure was the final barrier between the insight and the decision.

Each of these risks, on its own, can stall a good idea. But when two or three activate at the same time, the organizational immune system mobilizes fully.

The data gets questioned. The methodology gets challenged. The timeline gets extended. A subcommittee forms. The subcommittee creates a working group. The working group schedules a listening session. The listening session produces a summary document no one reads, except one person named Mark who loves process and has never once emotionally recovered from Robert’s Rules of Order.

This is how insight dies.

Administratively.

With calendar invites.

And to be clear, this isn’t always malice.

It’s not always stupidity.

Often, it’s self-preservation.

A strategy needs to change. A program needs to be protected. A leader needs feedback. A team already knows the answer but can’t afford the consequences of saying it plainly.

The program I described survived.

Barely.

It survived because someone in the room was willing to name what was actually happening.

The real disagreement wasn’t about criteria.

It was about cost.

The logic cost. The strategic cost. The social cost.

Once those costs were named, they could be managed.

Sometimes the work is the work you were hired to do.

And sometimes the real work is identifying the work that only becomes visible after the obvious work is finished.

That’s what happened here. The obvious work was evaluation. The real work was helping the room survive what the evaluation revealed.

Action

You don’t make insight move by making it more correct.

That sounds obvious until you’ve watched very smart people respond to resistance by adding more evidence to a room that already understands the evidence. They add another chart. Then a longer appendix. Then a cleaner summary. Then a pre-read with a subject line that says “for context,” which is usually where hope goes to stretch its hamstrings before pulling something.

The issue isn’t always that people don’t understand.

Sometimes they understand perfectly.

That’s the problem.

Because if they understand, then they also understand what accepting it would require.

You don’t make insight move by explaining it harder to people who already see the case. You make insight move by reducing the cost of accepting it in the room.

That starts with a better question.

Once the insight has been validated, the question isn’t only:

“Why are people resisting?”

The better question is:

“What will this insight cost, and who’ll be asked to pay?”

That question changes the work.

If the insight breaks the model people have been using to understand the work, you’re dealing with logic cost. If it disrupts a plan, timeline, investment, or public commitment already in motion, you’re dealing with strategic cost. If it makes someone look wrong, exposed, complicit, out of touch, or less competent, you’re dealing with social cost.

And if the resistance isn’t good-faith risk, but bad-faith obstruction, that has to be named too.

This matters because not all resistance deserves the same response.

Some resistance is intelligence. People aren’t rejecting the insight. They’re warning you that the proposed path will fail.

Some resistance is risk. People understand the insight, but they’re struggling to absorb its consequences.

And some resistance is obstruction. People use complexity, process, tone, timing, methodology, or endless requests for more proof to avoid accountability or protect power.

Those are not the same thing.

So they can’t be handled the same way.

This is where the work becomes practical.

When I help leadership teams work through this, I’m not trying to make the truth softer. I’m trying to make it survivable enough to move.

That might mean reframing the issue so leaders can update without unnecessary humiliation. It might mean redesigning the path because the insight is valid but the implementation plan will fail. It might mean resourcing the work, because unfunded priorities aren’t strategies. They’re vibes with deadlines.

In other cases, the work is repair because the insight reveals harm or broken trust. Sometimes it’s confrontation because process has become camouflage for avoidance. Sometimes the right move is to stop, because sunk-cost loyalty has become damage. And sometimes the wisest move is to pilot or escalate, because the insight is strong but the path still needs testing, or because the cost exceeds the authority of the current group.

That’s the intervention layer.

Not advice.

Not a better deck.

Not a motivational speech where everyone nods and then returns to the same incentives by Wednesday.

The point isn’t to make everyone comfortable.

The point is to stop pretending accuracy alone creates action.

It doesn’t.

Insight that people can’t afford to accept will not move.

No matter how true it is.

So think about the last decision in your organization that stalled. The initiative that lost momentum. The obvious answer that never became the chosen one. The thing everyone could see, but no one could safely say.

Was the insight wrong?

Or was the cost of accepting it too high?

If it was the second one, there’s a path forward.

The costs are nameable.

And named costs can be managed.

The work is not to look busier, sharper, or more productive.

It’s to find the insights that matter most, then make them survivable enough to change the organization.

Because smart organizations don’t fail only because they fear the truth.

They fail when intelligence becomes expensive and nobody knows how to pay the cost without humiliating the room.

Signal Boost

A line I keep returning to:

Data insight tells you what the evidence reveals. People insight tells you what the evidence threatens.

That distinction changed the way I think about strategy, assessment, AI implementation, culture work, leadership feedback, and organizational decision-making.

If you only have data insight, you may be right and still ineffective. If you only have people insight, you may be politically skilled and still untethered from reality.

The work is to hold both at the same time.

What the evidence reveals.

What the evidence threatens.

And what it will take for the room to survive accepting both.