File this under #management

I’ve spent the last few months experimenting with different product ideas. I’ve discarded most of them. But a select few show promise. This is one of them.

Drumroll

Lots of organizations collect and share 360 feedback. This is good. Feedback is a gift and should, as the “360” moniker implies, draw upon multiple points of view.

But feedback is personal and retrospective. It asks “reviewers” to pass judgment on someone’s past behaviour:

How did Sarah do in the last six months?
What did she do well?
What do you think she can do better?

Helpful, yes. But also a bit problematic.

Feedforward is different. It’s situational and forward-looking. It asks “advisors” for guidance on how to achieve a certain goal:

Sarah is on a mission to increase customer retention by 15% by year-end.
She’s managed 5% to date, with six months left to go.
What advice do you have to help her achieve her goal?

That’s a powerful shift that makes feedback better and more actionable. And so today, I’d like to introduce you to FeedFwd.app:

FeedFwd.app is a SaaS platform that helps managers at small to medium-sized organizations provide their people with continuous, 360-degree feedforward advice.

Here’s how it works:

1. The manager and employee agree on a mission the employee should achieve. They define the goal, the timeline, and a set of advisors to tap for advice.
2. FeedFwd automatically creates a schedule of spaced-out feedforward requests to ensure the employee receives advice throughout their mission.
3. Once the mission is underway, FeedFwd periodically requests status reports from the employee. These are shared with the manager and included in subsequent feedforward requests.
4. Advisors receive these requests by email. They’re asked to review status and provide advice on how the employee might ensure mission success.
5. All the while, FeedFwd handles requests and reminders automatically, sharing advice with both employee and manager as it arrives.

FeedFwd is built on the feedback practices we developed at MAQE. And it’s further shaped by our experience using various tools, ranging from manual tracking to purpose-built software.

In the end, the ability to continuously and automatically collect future-oriented, mission-focused advice felt like a use case compelling enough to warrant an MVP. And we’re about to open for public beta.

If you’re a manager with around ten direct reports who wants a better way to gather and distribute feedback, email me. You’ll get free access and personal support.

Cheers,

/Andreas

Elsewhere…

Don’t be an uncreative hack

Writes product leader Philip Carter in LLMs are Weird Computers: “[If] you’re a company leader who sees a wave as large as the introduction of the computer coming and your thought is to “use less resources to do the same work”, you’re an uncreative hack and it’s you who deserves to be fired. The goal should be how you can accomplish more when you have cognitive co-processors at your disposal.” Amen.

On the power of alignment

“In The Five Dysfunctions of a Team, Patrick Lencioni credits a friend who built an organization from a start-up to a billion dollars in revenue with the following observation: “If you could get all the people in an organization rowing in the same direction, you could dominate any industry, in any market, against any competition, at any time.” From Gino Wickman’s book Traction on the Entrepreneurial Operating System.

How can we build better organizations? That’s the question I’ve been trying to answer for the past 10 years. Each week, I share some of what I’ve learned in an email newsletter called WorkMatters. Back issues are marinated for three months before being published to Medium. This article was originally published on Jul 31, 2026. Subscribe to email updates at www.workmatters.studio.

Feedforward Turns Feedback Into Actionable Advice was originally published in WorkMatters on Medium, where people are continuing the conversation by highlighting and responding to this story.

File this under #organization

Some companies describe themselves as “mission-led”, others as “mission-driven”, yet others as “mission-controlled”. But what does this actually mean? Is there a difference?

I think yes. After reading Eric Ries’ new book Incorruptible, I’d define the terms as follows:

• Mission-Led Organizations have a pronounced purpose, mission, or “just cause” that advances human flourishing;
• Mission-Driven Organizations have a business model that allows them to profit only by advancing that mission; and
• Mission-Controlled Organizations have governance that protects that mission through changes in leadership and ownership, etcetera.

These are my definitions.
But they’re very much built on Ries’ ideas:

• Why human flourishing? Because purpose is morally neutral. Human flourishing distinguishes a worthy purpose or mission from a merely clear one.
• Why the economics tie-in? Because intentions are fragile when purpose and money conflict. Mission drive (Ries’ term) makes advancing the mission profitable.
• Why governance mechanisms? Because even aligned business models can change. Mission control protects the mission through changes in leadership, ownership, and financial pressure.

Five Levels

“Mission-led”, “mission-driven”, and “mission-controlled” are distinct but cumulative ideas. Together, they suggest a hierarchy of what we might call “mission alignment”.

Mission Alignment is the extent to which an organization’s mission, economics, and governance reinforce one another.

There are five levels:

1. Missionless: The organization has no purpose beyond survival or success.
2. Mission-stated: Has a mission without a commitment to human flourishing.
3. Flourishing-oriented: Has a mission that advances human flourishing.
4. Mission-driven: Profits only by advancing said mission.
5. Mission-controlled: Protects its mission through governance.

Again, this is my interpretation. But it’s close enough to Ries’ writing that I don’t think he’d object.

Each level answers progressively harder questions:

• Do we have a mission?
• Is it directed toward human flourishing?
• Does accomplishing it drive our economics?
• Will it remain in control when honoring it becomes difficult?

Becoming Incorruptible

Ries has seen too many organizations drift from their founding missions under the weight of success, market demands, and short-term financial pressure.

That’s why he wrote Incorruptible. The book offers a blueprint for building organizations whose missions can outlast their founders, making the mission itself the ultimate authority.

For my part, I hope the above five-level framework makes the progression a bit easier to see and think about. Perhaps it can even encourage you to level up?

/Andreas

Sidenotes

✊ Harder is Easier

There are two types of hard in business as in life. Writes Eric Ries in Incorruptible: “One hard is depleting. One hard is exhilarating. One hard makes talented people do mediocre work. The other calls forth their passion, creative thinking, and loyalty.” That’s worth keeping in mind — especially as it relates to missions to “advance human flourishing”. It sounds paradoxical, but a hard and ambitious mission might actually make it easier to find the right talent, partners, and customers.

❤️‍🔥 Reclaiming Purpose

Many leaders feel disconnected because their work lacks true meaning, writes Reinventing Organizations author Frederic Laloux. He’s come up with a three-question test to diagnose the problem: Ask yourself what would happen to your organization if (1) no marketing or advertising were allowed, (2) customers had to watch a 5-min video showing how your product or service was made, and (3) the price you charged reflected the full cost of production to society.

How can we build better organizations? That’s the question I’ve been trying to answer for the past 10 years. Each week, I share some of what I’ve learned in an email newsletter called WorkMatters. Back issues are marinated for three months before being published to Medium. This article was originally published on Jul 24, 2026. If you are reading this, you’re missing out. Subscribe to email updates at www.workmatters.studio.

The Five Levels of Mission Alignment was originally published in WorkMatters on Medium, where people are continuing the conversation by highlighting and responding to this story.

File this under #innovation

Remember the primer I shared a few years back on ISO56002, the standard for innovation management systems (see #149)? It’s since become a major focus of mine, and I am now building and consulting on the topic.

Governance is a key part of this work. Not (just) in terms of how to manage innovation work, but in terms of how to manage the system in which that work gets done. It’s a bit meta, but also very important.

Now, ISO standards are supposedly applicable to organisations of any size. But the reality is that users tend to skew large: think enterprise organisations with thousands of people spread across multiple SBUs.

Six Questions

Simplicity is a virtue at any scale, but doubly so here. Large organisations emphasise the “complex” in complex adaptive systems, which means that anything we can do to simplify things tends to be welcome.

I have, to this end, formulated six simple-but-not-simplistic questions for orgs to answer when tackling the governance questions:

1. Who sets the strategy and focus areas? Which problems are even in scope?
2. Who identifies opportunities? Where in the organisation do new concepts originate?
3. Who allocates funds? Money is the most concrete expression of governance power.
4. Who manages execution? Day-to-day project ownership, including who runs the stage gates.
5. Who measures outcomes? Who defines success, and how are results tracked?
6. Who manages the system? The innovation system needs an owner, or it drifts.

Note that short doesn’t mean easy. On the contrary, each question is likely to spark plenty of debate.

I have, however, found one thing that tends to make answering the above a whole lot easier: philosophical agreement on the right level of centralisation.

One Choice

Philosophy? In an article about innovation management? Damn, right. Because at the heart of the six questions lies a choice: What type of organisation do we want to be? One that values efficiency and control, or one that values creativity and speed?

Neither is necessarily better than the other. It all comes down to something I call Organisation-Market Fit (see #148), defined as the best possible organisational structure for whatever the prevailing market conditions might be.

My point is this: organisations that know (and agree) on who they are will have a much, much easier time answering the six questions. They’ll spend less time arguing and more time ironing out the details of their to-be governance structure.

Three Options

Speaking of governance structure: theory suggests there are three options to choose from: Centralised, Decentralised, and Hybrid/Federated:

• Centralised: Corporate sets the agenda, holds the budget, and oversees execution. Units implement but do not originate or decide.
• Decentralised: Business units own the full cycle, including their own innovation systems. The centre’s role is financial oversight, not innovation direction.
• Hybrid/Federated: Ownership is shared. The centre sets the framework and holds the portfolio view; business units originate ideas and run execution within guardrails.

This might be true. But I’ve yet to see an organisation that fits squarely in the Centralised or Decentralised camp. Virtually everyone picks a hybrid approach.

Why?

Because while the Centralized and Decentralized are useful for framing purposes, neither is actually feasible in real life.

Centralised governance would produce coherence and speed on big bets (good), while creating bottlenecks on operational decisions (bad).

Decentralised governance would produce speed and marketing relevance (good), but risk portfolio fragmentation and the quiet death of anything requiring a long time horizon (bad).

So the only real option is the Hybrid/Federated structure. And that’s where the Six Questions come into play. They’ll help you find the balance between strategic alignment and operational ownership that is right for you.

/Andreas

PS. For more on Organisation-Market Fit, see also my series on management spans starting with #302, #303, #304, and #305.

Elsewhere

👨‍🔬 INSEAD AI Venture Lab

I’ve joined the 10-week accelerator program hosted by INSEAD and Audos.com. I’ll be experimenting with a couple of different ideas, most notably something built atop the above-mentioned ISO framework.

😒 The Paradox of Creating

“When you are creating something, you will usually find that once things appear 90 percent done, you are actually about halfway there. But you have to create the first 90 percent to see the half that needs to be revised.” — James Clear

How can we build better organisations? That’s the question I’ve been trying to answer for the past 10 years. Each week, I share some of what I’ve learned in an email newsletter called WorkMatters. Back issues are marinated for three months before being published to Medium. This article was originally published on Jul 17, 2026. If you are reading this, you are missing out. Subscribe to email updates at www.workmatters.studio.

Six Questions on Innovation Governance was originally published in WorkMatters on Medium, where people are continuing the conversation by highlighting and responding to this story.

File this under #organization

This newsletter has, over the past weeks and months, argued that AI is changing how individuals work, how teams collaborate, and how organizations coordinate.

These are not isolated ideas. They are interrelated and mutually reinforcing, amounting to something of a flywheel for systemic change.

I’ve begun thinking of the destination as The Agentic Economy — an economic era brought on by AI and powered by new economics of coordination.

Here’s what it might look like:

Individuals do less, decide more. Execution gets cheap. Judgment doesn’t. The valuable people know what’s worth producing, not just how to produce it fast.

High agency becomes a must-have. Knowing what should be done before anyone asks you to do it. That’s the trait AI can’t replace, and increasingly the one that matters most.

Autonomy moves up the ladder. Micromanaged, managed, self-managed, self-directed, self-governed — as agency rises, people climb it. Fewer roles stay stuck at the bottom.

Teams shrink. A handful of generalists with a portfolio of agents can do what used to take a department. Headcount stops being the answer to “we need more capacity.”

Teams form around the work, not the org chart. People assemble around a specific opportunity, then dissolve once it’s done. The team is the unit, not the function.

Managers stop routing and start designing. Hierarchy was always partly an information system. What’s left for managers is designing that system, not running it.

Org AI takes over the routing. Something tracks every project, decision, and conversation, and makes sure people have what they need. No manager required in the loop.

Organizations decentralize. With Org AI handling coordination, small autonomous teams run without layers of management holding them together.

The big gets bigger. Decentralization notwithstanding, scale still wins on distribution, brand, and trust. That means fewer, larger platform companies within which small teams operate.

People and companies look more like partners. Teams carry real ownership of outcomes. The relationship starts to look less like employment, more like a deal between two parties.

People stop climbing ladders. Range, reputation, and the ability to lead a scenario matter more than a title. “Climbing the ladder” means less when structures are flat.

Labour gets more mobile. If teams form and reform around scenarios, and orgs are networks without solid walls, people can move in and out of them more freely.

…

Electricity brought on the Industrial Economy. Computers and the Internet brought the Information Economy. And now AI might push us into the Agentic Economy.

/Andreas

PS. The term “agentic” in the title does two things: it acknowledges AI’s role in this change, and it underscores the talent profile at the centre of it.

Elsewhere…

💫 Building a Self-Improving Company

Tom Blomfield, Monzo co-founder and GP at Y Combinator, likens an Organizational AI to a control system that continuously senses reality, acts autonomously, evaluates continuously, and learns and adapts its own behaviour, all with minimal human intervention. “Think about each part of your company,” he says, “as a self-improving, recursive AI loop.”

How can we build better organizations? That’s the question I’ve been trying to answer for the past 10 years. Each week, I share some of what I’ve learned in an email newsletter called WorkMatters. Back issues are marinated for three months before being published to Medium. This article was originally published on Jul 10, 2026. If you are reading this, you’re missing out. Subscribe to email updates at www.workmatters.studio.

The Agentic Economy was originally published in WorkMatters on Medium, where people are continuing the conversation by highlighting and responding to this story.

File this under #management

Last week’s newsletter promised an exploration of the second-order effects of AI adoption. The potential effects, I should say. No one really knows what those effects might be. But we can speculate. And so with history as a guide, that’s exactly what we are going to do.

But first:

Last week’s story on electrification provided us with a two-part analogy: the drive shaft as the factory’s circulatory system, and unit drive as the innovation that transformed said system from central to distributed control.

The analogy is a good one. So good, in fact, that we’ll use it as the jumping-off point for our exploration. And the first step in that endeavour is to find AI-age equivalents for both drive shafts and unit drive.

As luck would have it, I think I’ve got strong contenders for both:

The hierarchy is the drive shaft

Ok, so the drive shaft was the factory’s circulatory system. It carried power from a single engine to multiple machines. It was a centralized system that dictated most everything about how work got done.

Today, the organisational hierarchy does much the same thing. Sure, it’s a different kind of circulatory system, carrying a different kind of power (information). But it, too, is centralized, and it, too, dictates most everything about how work gets done.

Consider the following:

At the turn of the century, factories had to be built tall and narrow. The drive shaft had limited reach, which meant machines had to be arranged in rows and across floors.

Today, organizations, too, are built tall and narrow. Management span of control is now the limiting factor, which means hierarchies grow ten or more levels deep.

As equivalents go, it’s a good one.

Organisational AI is the unit drive

Unit drive allowed individual machines to be fitted with their own power source. It was a paradigm shift. But not because the technology was new (the electric motors already existed), but because it turned a centralised circulatory system into a distributed one.

Organisational AI could do the same to the hierarchy.

Organizational AI is the use of AI as an organisational capability, not just as individual tools. It means embedding AI into the way an organisation senses information, makes decisions, coordinates work, learns from outcomes, and improves its operating system.

Today, the organization’s circulatory system is made up of people. Managers at each level decide what information moves when and where. But managers are not always available. They don’t always know what to share and to whom. And because information is power, they sometimes don’t want to share at all.

The result is information asymmetry. Some people get what they need. Others don’t.

Organisational AI could change this. Not by simply routing better, but by removing the need to route at all. Instead of information being pushed down the hierarchy, with judgment calls being made at each level, it’s pulled when and where needed, 24/7, 365 days per year.

At that point, the manager stops being the valve. And once no one has to route information for a living, the hierarchy loses much of its raison d’être.

The future is distributed

So what about those second-order effects?

Well, if history is any indication, we could be looking at the end of hierarchy. Or, rather, the beginning of the end.

Yes, Organizational AI could do to the hierarchy what unit drive did to the drive shaft: make it superfluous and, in so doing, enable a distributed system to supplant a centralized one. But remember the assertion at the top of last week’s newsletter:

The current state of AI is akin to the early days of electrification.

Early days indeed.

First, the technology arrives, and we put it to work inside the existing setup. The electric motor replaced the steam engine; AI arrives as copilots. The effect is immediate but also modest: the same work, only cheaper, faster, and more efficient.

That’s where we are now.

Then the real innovation follows. Unit drive in the factory; Organisational AI in the firm. The effect is larger but also more long-term: the circulatory system gradually shifts from central to distributed.

That’s what’s next.

Then, last, our ways of working are rebuilt around this newfound freedom. In the factory, it gave rise to innovations like the assembly line. In the organisation, we’re looking at new and better ways of organizing, the details of which are still unclear.

But we do know one thing: this will take time. Systemic change always does.

Unit drive took years to reach critical mass. And once it did, it took a long time before anyone realized just how superior the distributed factory actually was.

The same is probably true of Organisational AI.

These systems are not yet widely available. But they’re coming. And only once they reach critical mass will their second-order effects start to show.

/Andreas

Elsewhere

😔 The long shadow of bullying

A new episode of Heavyweight tells the story of Julia, a 35-year-old New York Times journalist who still lives in fear of her middle-school bullies. It’s an important reminder that bullying isn’t just about childhood cruelty; for some, it becomes a trauma that follows them well into adulthood. Heavyweight has, btw, become one of my favorite podcasts.

How can we build better organizations? That’s the question I’ve been trying to answer for the past 10 years. Each week, I share some of what I’ve learned in an email newsletter called WorkMatters. Back issues are marinated for three months before being published to Medium. This article was originally published on Jul 3, 2026. If you are reading this, you’re missing out. Subscribe to email updates at www.workmatters.studio.

The Hierarchy Is the Driveshaft was originally published in WorkMatters on Medium, where people are continuing the conversation by highlighting and responding to this story.

File this under #organization

The current state of AI is akin to the early days of electrification. The technology is here and is rapidly spreading, but we’ve yet to realize its full potential.

Case in point: In their State of AI report, McKinsey reports that 94% of companies they’ve surveyed have yet to see significant financial value from their AI investments.

The analogy is a good one, and I’m not the first to make it. But there’s more to this story than what first meets the eye.

Before the motor

Before the 1890s, factories had but one power source: the water wheel, later the steam engine. Power was generated in one place and distributed elsewhere through steel shafts.

These so-called drive shafts would typically run in or near the ceiling, and they would connect to smaller, vertical belts that connected the central drive shaft to individual workstations.

This arrangement decided everything about how work got done. Machines had to be lined up in rows, their placement dictated not by workflow but by drive shaft proximity.

Factories were built tall and narrow for the same reason. Drive shafts could not travel far, which meant work had to be spread across floors and up and down stairs.

It was less than ideal. And dangerous. Power was either on or off everywhere all at once. And the exposed mechanics meant injuries were common.

A false start

When the electrical motor arrived, it changed almost nothing. Not right away.

Steam engines were replaced by their electrical counterparts, but the exposed drive shafts remained. And factories continued to be built tall and narrow.

The technology was there, but the understanding of how to make good use of it remained stubbornly elusive.

Adoption lagged. Come the turn of the century, less than 5 percent of American factory machinery was powered by electricity. And most of that still through a central drive shaft.

Freedom from constraints

The breakthrough came in the form of unit drive when individual machines were equipped with their own electric motors.

This might seem like a simple enough innovation, but let’s remember that the second- and third-order benefits of that change were completely unknown. We just didn’t know.

Sure, replacing a single motor with unit drive would allow the factory to remove its drive shafts. But why? Retrofitting would be expensive, and the drive shafts already worked. So why rock the boat?

The fact that this “small” change would allow us to redesign both factory and workflow, and that this redesign would unlock productivity gains previously unthinkable was… well, unthinkable.

And that’s where we stand in terms of AI today.

We have the technology. But we’ve barely begun to use it. And we don’t yet know what the Nth order consequences might be.

And that’s where we’ll pick things up next week.

/Andreas

Elsewhere…

🏃‍♂️‍➡️ Runner’s High

Last month, Rachel Entrekin ran the 400km Cocodona250 ultramarathon in 56 hours and 9 minutes. She won the race outright, beating the course record. And as impressive as that sounds, I’m even more impressed by her attitude. As she explains in this Rick Roll interview, her only goal in entering the race was to enjoy the experience and practice the person she wanted to be under pressure.

How can we build better organizations? That’s the question I’ve been trying to answer for the past 10 years. Each week, I share some of what I’ve learned in an email newsletter called WorkMatters. Back issues are marinated for three months before being published to Medium. This article was originally published on Jun 26, 2026. If you are reading this, you’re missing out. Subscribe to email updates at www.workmatters.studio.

Freedom From the Driveshaft was originally published in WorkMatters on Medium, where people are continuing the conversation by highlighting and responding to this story.

File this under #fulfillment

It’s high time for another readership survey.

I’ve been writing WorkMatters for more than 8 years at this point. And I have, during that time, covered a lot of different topics: from personal productivity and product development on the one hand, to leadership practices and organisational design on the other.

I’ve also experimented with a lot of different formats — from short practical explainers on things like OKRs, PLG, and ISO 56002, et cetera, to series of long-form articles on big thorny topics such as economic engagement, self-organisation, and artificial intelligence.

This isn’t a surprise. Things change. I know I’ve changed. I used to run a smaller company. I then scaled that company to 100+ employees. I then lost everything. And now I run a small company again, all the while experimenting with new product ideas.

A bit of strategic drift is to be expected.

But here’s the thing: as a WorkMatters reader, you came here for a particular reason. And whatever that reason was, you’ve since been the recipient of a somewhat eclectic mix of topics best described as whatever was on my mind at the time.

Understandable, yes.
But not ideal in terms of content strategy.

I had a few different goals in mind when I began writing in 2018. I wanted to remember more of what I read, and I knew writing was a great way to do that. I also wanted to teach myself to ship, which is why I forced myself to publish weekly. Fridays at 10 AM ICT, rain or shine.

The goal of sharing and learning from others came later, but it’s now a primary reason why I do what I do. And this is why the “strategic drift” bothers me. I want WorkMatters to be useful to you, and that’s why this survey is important.

Take Survey

It’ll only take a few minutes, and your feedback is much appreciated.

Cheers,

/Andreas

Elsewhere…

📕 Gino Wickman on getting traction

I just finished Traction — Gino Wickman’s intro to the Entrepreneurial Operating System (EOS). It’s an easy read that I’d recommend to anyone running a growing company. It provides a collection of best practices that, when combined, creates a complete operating system.

🎧 Benedict Evans on AI in proportion

Technology analyst Benedict Evans thinks that AI is as big of a deal as the Internet and mobile, and only as big a deal as Internet and mobile. Life and work are going to change in ways we cannot predict. But we’ve been here before. We’ll probably be OK.

How can we build better organizations? That’s the question I’ve been trying to answer for the past 10 years. Each week, I share some of what I’ve learned in an email newsletter called WorkMatters. Back issues are marinated for three months before being published to Medium. This article was originally published on Jun 19, 2026. If you are reading this, you are missing out. Subscribe to email updates at www.workmatters.studio.

Readership Survey 2026 was originally published in WorkMatters on Medium, where people are continuing the conversation by highlighting and responding to this story.

File this under #innovation

How do you know if your organization is innovative? A team at McKinsey decided to find out. Their work began in the early 2010s, was first published in the 2015 article The Eight Essentials of Innovation, and was later validated in a 2019 follow-up study.

Here’s what you need to know:

The Research

The original study took several years to complete and covered interviews, workshops, and surveys with 2,500 executives across 300-plus companies of various industries and locales. In addition to questions about discrete innovation practices, the study also asked about success metrics — data that was later used to distinguish the top innovators.

The Model

Armed with this information, the team went on to identify 30-or-so practices that set the top innovators apart. And from these, they distilled eight essential questions — four strategic (e.g., Aspire, Choose, Discover, Evolve), and four organizational (i.e., Accelerate, Scale, Extend, and Mobilise) — that organizations can use to assess themselves:

Source: The 8 Essentials of Innovation (2015)

The Findings

The original study’s claim was that top-innovators had or did the above essentials with greater consistency than their less innovative counterparts. But a subsequent 2019 follow-on study went further: using a smaller set of ~180 companies, they were able to show that companies that mastered at least five of the eight essentials performed 2.4x better financially.

…

So, how do you know if your organization is innovative?

Well, supposedly you start by answering the above 8 questions. You know you are innovative if you score five or above.

;-)

A few caveats to keep in mind:

• The original study was largely self-reported. And even with the follow-up study, it looked at correlation, not causation. As Claude told me during prep: “It’s an open question whether the essentials produce success or if success affords the essentials.”
• An 8-question diagnostic can only take you so far. And that’s by design. The practices and competencies that underscore the 8 essentials are all part of McKinsey’s proprietary data set. A more comprehensive diagnostic is available for customers.
• The essentials is about attributes rather than mechanisms. It tells you what to do, not how to do it. McKinsey is of course happy to help, but you might also look at ISO 56002 — the guidance standard for innovation management systems.

That’s all for this week.
Until next time, make it matter.

/Andreas

Marginalia

Two recommendations this week from Lenny’s Podcast:

✊ Incorruptible

Eric Ries, author of Lean Startup and founder of the Long-Term Stock Exchange, has seen too many mission-driven founders get pushed out of the companies they’ve built. His new book — Incorruptible — explains why it happens and how to prevent it.

✋ Irreplaceable

Dan Shipper, co-founder at Every, doesn’t believe in the AI job apocalypse. He says that more automation means more work, and that the people who thrive will be the ones who learn to direct the machines rather than compete with them.

How can we build better organizations? That’s the question I’ve been trying to answer for the past 10 years. Each week, I share some of what I’ve learned in an email newsletter called WorkMatters. Back issues are marinated for three months before being published to Medium. This article was originally published on Jun 6, 2026. If you are reading this, you’re missing out. Subscribe to email updates at www.workmatters.studio.

🌱 The 8 Essentials of Innovation: Research, Model, and Findings was originally published in WorkMatters on Medium, where people are continuing the conversation by highlighting and responding to this story.

File this under #management

There is a debate in Silicon Valley right now about how leaders should behave. Founder Mode: stay close to the work, bypass layers, maintain direct contact with what actually happens; or Manager Mode: delegate, trust the system, manage from a distance.

It is an interesting conversation, but also a narrow one. Both modes assume someone is firmly in charge. The question is only how they exercise that authority.

WorkMatters has been interested in a different question: whether organisations can be structured so that centralised, permanent authority matters less. Not leaderlessness, necessarily; someone is always in charge, and that is fine. The question is whether it always has to be the same person. Self-organising teams distribute authority to whoever is best positioned to make a given decision. Leadership becomes contextual rather than positional.

These two conversations have felt largely incompatible. Then I listened to Ivan Zhao, co-founder and CEO of Notion, on Brian Halligan’s Strange Long Trip podcast.

Most companies start small. The founder is in the room, trust is high, communication is direct. No one needs a detailed plan because everyone knows the tune. Then the company grows, layers appear, and the instinct, rational enough, is to write things down and coordinate from the top. The jazz band becomes a marching band. Everyone plays their part on cue, following a score written by someone else. Stable, but slow. Coordinated, but initiative-poor.

Zhao calls the alternative Jazz Mode. The jazz band still has a leader: someone who sets the tune, the key, and the tempo. What it doesn’t have is a score everyone follows note for note. Within that structure, musicians listen, respond, and improvise. And that’s important, given where we’re at. With AI, every week brings something new that wasn’t possible the week before. Under those conditions, the marching band has a serious problem that the jazz band doesn’t.

To be clear, Jazz Mode is not self-organisation. But it is not Founder Mode or Manager Mode either. Those debates are about who holds authority. Jazz Mode is about how authority flows. It is maybe the most honest starting point I have come across for anyone trying to build something that doesn’t eventually collapse into a marching band.

/Andreas

Sidenotes

😎 Agency

Zhao mentions that Notion has shifted what it hires for: less experience, more agency and taste. Language models have democratised capability. Almost anyone can write and code adequately now. What they cannot replicate is judgement about what is worth building, and the will to go build it.

🌱 Refounding

If your organisation feels more like a marching band than a jazz band, the instinct might be to pivot. But a pivot changes what you do. Zhao’s word for what you actually need is a refounding: it changes who you are. He has done it twice at Notion, and each time the question was identity rather than strategy.

How can we build better organizations? That’s the question I’ve been trying to answer for the past 10 years. Each week, I share some of what I’ve learned in an email newsletter called WorkMatters. Back issues are marinated for three months before being published to Medium. This article was originally published on May 29, 2026. If you are reading this you are missing out. Subscribe to email updates at www.workmatters.studio.

Learn to Play Organizational Jazz was originally published in WorkMatters on Medium, where people are continuing the conversation by highlighting and responding to this story.

File this under #organization

Something interesting is happening. Two worlds that rarely see eye to eye are quietly converging on the same idea.

The first is the tech world: founder mode, venture capital, and hyperscale. The second is what some call Newwork, a looser movement built around self-management, collaboration, and purpose. On most things, they occupy opposite ends of the spectrum.

But not on managers. Tech finds them problematic; too many layers slow the organization down and dilute the founder’s vision. Newwork finds them superfluous; given the right conditions, people can manage themselves. Both worlds have been waiting for something to make managers optional.

They may have found it.

The Factory Problem

Research suggests that while enterprises are investing heavily in AI, few are seeing much return. “Individual AI” is a genuine boon to productivity, but those gains tend to get absorbed by the organization at large. We have yet to solve the “Enterprise AI” question.

The analogy to industrialization helps. Factories that introduced machinery quickly saw productivity spike in specific areas. But could not realize the full potential of their new tools until they redesigned the factory itself. The coordination layer in between was the bottleneck.

The same logic applies now. With AI serving as an intelligence and coordination layer, organizations could share information faster, push decisions to the front lines, and do much of what managers currently do, more consistently and at lower cost.

At which point the question becomes an uncomfortable one: what exactly are managers for?

Control or Liberation

This idea appeals to both worlds, albeit for different reasons.

In tech circles, fewer management layers means the founder, enabled by AI, can more easily enforce their vision throughout the organization. This is “Founder Mode” in practice: deeply involved leadership, with AI handling the coordination work that would otherwise require a management layer to execute.

In Newwork, the same intelligence layer looks entirely different.

Here, it’s not a tool for top-down control. It’s a potential equalizer, one that ensures everyone has access to the same information at the same time. Rather than amplifying the founder’s reach, it could flatten hierarchy by removing the information asymmetry that hierarchies depend on and perpetuate. For Newwork proponents, this is the moment self-management has been waiting for.

The Work That Remains

Neither world is likely to get exactly what it wants.

For tech, AI will compress rather than eliminate management. Fewer layers, faster information flow, more direct alignment from the top. But even Founder Mode needs judgment and trust at the edges.

For Newwork, AI may remove some of the coordination burden that made self-management hard to scale. But self-organization still requires structure; someone still has to clarify expectations, resolve tensions, and help people make good decisions together.

What AI changes is not whether management exists, but what it is for.

If AI absorbs more of the reporting, reminders, documentation, handoffs, and information routing, management becomes less about keeping the machine running and more about improving the people and the system.

/Andreas

Sidenotes

💪 Paul Graham on Founder Mode

In a widely-read 2024 essay, Paul Graham argued that founders should run companies differently from professional managers. Where manager mode means hiring good people and stepping back, Founder Mode means staying deeply involved: skipping levels, talking directly to the people doing the work.

🦾 Brian Chesky on AI Founder Mode

Airbnb’s Brian Chesky is Founder Mode’s most cited practitioner. In a recent interview, he argued that AI makes the approach even more intense: fewer layers, a shift from meeting-based to asynchronous culture, and everyone expected to do work, not just manage it. His benchmark: the Catholic Church has run for 2,000 years on four layers.

🧠 Jack Dorsey on the Intelligence Layer

In a recent interview, Block’s Jack Dorsey named his biggest business regret: delegating too much. His answer is to rebuild around an intelligence layer that lets the founder’s intent flow directly through the organization, without holding company structures or management layers filtering the signal. AI, in his telling, makes Founder Mode scalable.

👍 Corporate Rebels on Role Clarity

Pim de Morree, co-founder of Corporate Rebels and a leading Newwork voice, recently connected Claude Code to Nestr.io, the platform his team uses to manage work roles, and handed his financial reporting role to AI entirely. The reason it worked: in self-managing organizations, roles are already clearly defined. That clarity is exactly what AI needs.

How can we build better organizations? That’s the question I’ve been trying to answer for the past 10 years. Each week, I share some of what I’ve learned in an email newsletter called WorkMatters. Back issues are marinated for three months before being published to Medium. This article was originally published on May 22, 2026. If you are reading this your are missing out. Subscribe to email updates at www.workmatters.studio.

Management Isn’t Dying. Coordination Is Being Automated. was originally published in WorkMatters on Medium, where people are continuing the conversation by highlighting and responding to this story.