BONUS Why Every Organization Reinvents Silos—And What to Do About It With Roland Flemm

BONUS: Why Every Organization Reinvents Silos—And What to Do About It

Today we speak with Roland Flemm, co-creator of Org Topologies and co-author of 10X Org — Powered by Org Topologies. Roland has spent decades in the trenches—first as a developer, then in infrastructure, and finally as a Scrum Master, trainer, and organizational design consultant. In this episode, he explains why even teenagers with zero corporate experience instinctively create departmental silos, why making every team faster doesn't make the whole organization faster, and how leaders can use the Org Topologies map to see their organization as it actually is—not as the org chart says it should be.

From Developer to Org Designer: Four Decades of Hitting the Same Wall

"I felt many, many times the limitations of organizational structures stopping me from using my common sense to make people work together in a proper way."

 

Roland's career spans over 40 years, starting as a developer in 1984. After a decade writing code and another decade in infrastructure, he moved into Scrum and agile coaching. But even as a highly effective Scrum Master, he kept hitting the same ceiling: local team improvements couldn't break through organizational boundaries. You could have wins with your team, but the moment you needed multiple teams to work together, someone higher up would shut it down. That frustration led him to Large-Scale Scrum (LeSS) by Bas Vodde and Craig Larman, which offered a more educated approach to multi-team collaboration—and eventually to co-creating Org Topologies as a way to help leaders see and change the structures that block real collaboration.

Bas has been on the podcast to share his view on scaling Scrum with LeSS, listen to his episode here.

The Hydrogen Car That Built Its Own Silos

"If you don't think about your org design—the way that you want to collaborate—then something like this happens."

 

One of the most striking stories in Roland's book comes from the Technical University of Delft, where student engineers were thrown together to build a hydrogen racing car. These were teenagers—no corporate experience, no boss who'd worked in a traditional company. And within weeks, they'd organized themselves into departmental silos, each sticking to their specialty. The mechanical engineers stayed on their turf, the electrical engineers on theirs. It was automatic. Roland traces this instinct deep: from school, where you choose a specialty; from the army and the church, where hierarchy is the default; from society itself, where "you're a plumber, so then we know what you are." The pattern of drawing boundaries and appointing leads when faced with complexity isn't corporate culture—it's human nature. And the problem isn't that it exists. The problem is that we don't know there are alternatives.

The Ferrari Effect: Why Local Speed Creates Global Congestion

"It's not that people choose to do fewer things. They just push more into the system because it can handle it. And that's where things go wrong."

 

Roland uses a vivid analogy from the book: swapping every car on the road for a Ferrari doesn't fix traffic congestion. The same principle applies in organizations. Everyone feels faster individually—teams are delivering, sprints are moving—but the whole isn't getting better. The HealthCare.gov story makes the case dramatically: 55 vendor firms, $1.7 billion in spending, and on launch day, six people successfully enrolled. Then a ten-person cross-functional team fixed it in six weeks. Roland sees this pattern repeat in banks that adopt delivery-oriented structures like SAFe: they create value streams, but because they don't make hard choices about what not to do, the freed-up coordination capacity immediately fills with new demands. The congestion returns, just at a different level.

In this segment, we talk about the Cynefin Framework. 

Three Topologies: Resource, Delivery, and Adaptive

"The third topology is interesting—that's where the hands and the heads are merged. They're no longer separated."

 

Roland walks through the Org Topologies map, each suited to different contexts:

 

Resource Topology — The "hands" are separated from the "heads." Coordinators design and direct; specialists execute narrow, deep tasks. This works in environments with low variability and deep technical expertise—think ASML's university-level hardware engineers, or a bank's core transaction processing team running COBOL. The focus is on utilization of expensive specialists.

 

Delivery Topology — Still has coordination overhead, but teams are cross-functional and can handle more complex problems end to end. A team owns the customer page and does design, testing, and deployment. This model favors speed of delivery, but breaks down when new work doesn't fit neatly onto existing value streams—like needing a retention initiative when no retention team exists. Work falls through the cracks.

 

Adaptive Topology — The hands and heads merge. People who coordinate can also do the work, and they self-organize around problems as they emerge. It's like a startup—"four guys and a dog in a garage"—but with hundreds of people. This model thrives in high-variability, high-learning environments where the investment in cross-training pays off because the challenges keep changing.

 

The key insight: none of these is "better." It's about fit for purpose. A single organization—like a large bank—might need all three topologies operating simultaneously in different parts of the business.

The MADE Loop: Map, Assess, Design, Elevate

"First, we all agree that the system that we're looking at is really the system that we're looking at. And then we can start talking about how to improve."

 

Rather than the typical transformation playbook—hire consultants, roll out a framework, hope for the best—Roland advocates for the MADE loop: Map the reality of how work actually flows (not what the org chart says), Assess whether that structure is fit for the strategic purpose, Design targeted improvements using the Org Topologies map, and Elevate through small experiments. Maybe two teams temporarily share members. Maybe one person switches team membership for a sprint. The changes are gradual, measurable, and reversible. Roland is emphatic about one principle from the book: "Own, Not Rent." Real structural change can't be outsourced to a consulting firm. Leaders have to see the system themselves—go to where the work happens, understand the flow, and make informed choices about what to change.

AI Is About to Reshape the Map

"As AI comes, you might want to get at least a part of that work transferred lower in the organization to more execution-oriented teams, because they can now use resources like AI to make proper decisions."

 

Roland makes a forward-looking point about how AI will shift the boundaries between topologies. Work that required deep specialist silos—like legal review or compliance decisions—may soon be handleable by cross-functional teams using AI tools. This means the threshold for when an adaptive or delivery topology makes sense will shift. Organizations that understand their current topology will be better positioned to adapt; those that don't will find their structures obsolete without understanding why.

 

About Roland Flemm

 

Roland Flemm is co-creator of Org Topologies and co-author of 10X Org — Powered by Org Topologies (2026) — a framework and book about elevating organizational performance through people-centered, strategy-driven redesign. He works with leaders in scale-ups and enterprises across Europe, helping them see how their org structure shapes — or blocks — their ability to learn, adapt, and deliver.

 

You can link with Roland Flemm on LinkedIn.

Learn more about Roland’s work at 10xorg and https://www.orgtopologies.com