Agencies are built to adapt. It's their superpower—and their curse. The same instincts that make them brilliant under pressure also make them nearly impossible to change. They are creative, fast, deadline-driven organisms that thrive on improvisation and heroics. And that's precisely the problem.
When you've built a culture of extraordinary adaptation, every new process looks like an interruption, and every structural improvement feels like a threat to the very muscles that make the place run.
Over the last decade, I've seen what happens when the adaptation habit meets a truly better way of working. Pods, when implemented fully, make agencies profoundly better. They cut meetings by 40 percent or more, shorten delivery cycles, reduce rework, and give people their evenings back. We have clients now celebrating their 260th sprint—ten full years of sustained improvement. The evidence isn't theoretical; it's sitting in the data, week after week, showing higher output, calmer teams, and better margins.
And yet, even in very large, sophisticated agencies, getting pods to take hold can be extraordinarily difficult. I'm working with one such agency right now—a global organization trying to launch a 45-person pod. Despite the compelling proof and strong intentions, consensus has been elusive. Many of the most effective aspects of the model have been stripped away in compromise, sacrificed in the name of "getting something started." The result is a structure that only vaguely resembles a true pod—and the benefits, predictably, are diluted.
Here's the paradox: it is precisely because agencies are so amazing—highly adaptive, deadline-driven, and capable of creative brilliance under chaotic conditions—that they are so difficult to change. Their adaptive genius, the same force that lets them thrive under pressure, also makes them resistant to reform. The more an organization depends on heroic adaptation, the less space there is for systemic improvement.
The Necessary but Uncomfortable Truth
In our leadership-foundation workshops—which are mandatory precisely because of what I'm about to say—I often begin with this line: "The biggest impediment to change is in this room."
The obstacle isn't on the front lines. It's in the middle: the discipline leads, the client managers, the project managers, the department heads. The very people who most want things to be better are the ones most likely to stall the change. That's not an accusation; it's an observation.
In Unmanaged, I wrote that "the ethos of management in your organization will have a profound effect on managers' choices." Managers who have learned to survive chaos develop reflexes that protect it. In over-matrixed agencies, that chaos is the native climate. And, as I note elsewhere in the book, truly effective structural or process change almost always requires a top-down mandate.
When that mandate is missing, as in this case, the best intentions of the middle layer tend to dissolve into negotiation and compromise—which is exactly how good ideas quietly die. Without a mandate, "must" doesn't exist. The elements of change become just another set of optional components in the already-crowded basket of agency processes. Everyone can pick and choose.
An agency can successfully install a new workflow management system because leadership mandates it—people may grumble, but nobody wants to propose a competing system. It's easier to comply than to rebel. But when the change being introduced is adaptive—a team-empowered structure like pods—it ironically invites everyone to throw in their own two cents about how they think it might work better. That reflex to customize, debate, and tweak may look collaborative, but it often derails the change before it begins.
The irony is painful: the system we're proposing has been refined over more than a decade, across more than 100 agencies, yet every new organization feels compelled to "personalize" it before even trying it as designed.
The Adaptive Fortress: A Structure and an Attitude
The middle layer has learned to survive precisely because the system rewards survival. Agencies prize deadline-driven results over healthy process. They celebrate the person who finds a way, even if that means inventing work-arounds and living on adrenaline. Over time, these behaviors harden into what I call adaptive fortresses—personal systems of meetings, relationships, and informal routines that let people deliver despite the dysfunction.
And agencies quietly reward that. The people who can keep projects alive in a broken environment rise fastest. Their skill isn't in process…it's in coping.
The adaptive fortress isn't just a structure; it's an attitude, a belief that real change is naïve, and that true professionalism is the ability to survive without it. That's why, when leadership introduces something like pods, it doesn't sound like freedom; it sounds like an existential threat. A structure designed to make work sane threatens the very coping mechanisms that have defined success.
Why "Better" Sounds Dangerous
Most of these leaders have been through a half-dozen (often many more) "better ways" before—Agile, squads, OKRs, collaboration tools, "new operating models." Each arrived with fanfare, each produced a flurry of meetings and metrics, and each faded quietly once the novelty wore off. They aren't resisting improvement; they're resisting heartbreak. They've seen "innovation by initiative" enough times to know the pattern.
As Unmanaged warns, "Even one or two Angry Ranchers can undo the good managing of many Humble Gardeners." When a few managers defend their personal systems, the rest learn to wait it out—and the organization slides back into its familiar rhythm of firefighting and heroics.
The Hidden Missing Ingredient: Shared Meaning
Most change programs fail before they start because people don't share a clear sense of why the change matters or how success will be judged. The metrics are fuzzy, hidden, or inconsistent. In Unmanaged, the very first of the Four Moments is the Why Moment—the point where everyone should understand the reasons for the work. But in most agencies that moment never really happens. Leaders skip straight to the What ("We're launching pods") without the common frame of why and how we'll measure better.
Metrics, when used well, are not control devices; they're meaning devices. Metrics tell people how improvement will be recognized. Without that clarity, every change feels arbitrary—another promise of "better" with no proof that it will be judged fairly.
From Persuasion to Cultural-Linguistic Change
This is where the standard agency change playbook fails. You can't persuade your way through adaptive fortresses. You have to change the language inside them.
Drawing on David Prochaska's model, real change begins in the pre-contemplative stage: when people are still talking about the change, testing ideas, and creating a shared vocabulary for what "better" even means. That's what I call building a cultural-linguistic narrative: a richer, collective story about the future that people can actually inhabit.
If you skip this stage (by treating pods as a structural fix instead of a cultural-linguistic shift) you trigger the same defensive reflex as every prior initiative. But if you engage the middle layer early in a larger conversation, and if you let them help name the change, define the metrics, and describe what success will feel like...you begin to dissolve the fortress from the inside. People start hearing themselves in the story. That's when the language of resistance turns into the language of ownership.
The Paradox of Improvement
The paradox is that the best way to help these managers is to make their world quieter, but they can't hear that promise until the noise is gone. So you have to show them, not tell them. Start with a single pod, supported by clear metrics and a common language for success. Let people feel what calm, focus, and flow actually feel like.
As Unmanaged points out, "Fully informed workers and teams get things done quickly and well, and they enjoy doing it." Once managers experience that calm—when they see projects running faster with less oversight—the argument for pods (and the needed changes) becomes self-evident. Metrics then serve as both mirror and proof: throughput up, meeting load down, satisfaction rising. Numbers tell the story that words alone can't.
How Change Really Spreads
The beauty of this Cultural-linguistic change is that it travels through conversation, not decree. When people start repeating the new language — "flow," "alignment," "managerial tax," "pods," "calm productivity" — you know the shift has begun.
That's why we make leadership workshops mandatory: the new language has to start at the top and be spoken consistently downward. Otherwise, old vocabulary rushes back to fill the silence. Change isn't enforced; it's narrated. And that narrative has to be bigger than the structure—it has to describe the better life that structure enables.
Closing Thought
Good ideas don't fail because they're unproven; they fail because they collide with the ingenious survival systems people have built to live inside bad ones. To break through, leaders must do three things at once when initiating change:
- Build shared meaning. Clarify the why and the metrics that define "better."
- Create a cultural-linguistic narrative that people can join and reshape.
- Prove the calm. Show, in a contained experiment, that the new system really does make life saner and work smoother.
Do those three, and the middle layer stops defending its fortresses and starts building bridges.
Postscript: The Tragedy of Partial Adoption
What makes this particular case especially sobering is that the agency has, in fact, embraced pods wholeheartedly. Leadership understands the model and its potential to calm the chaos of multi-client, multi-project life. They've begun introducing pods across several major accounts.
But the one pod chosen as the proving ground—the one that will inevitably become the template for all the others—has been compromised by habit. In practice, it runs like business as usual. The meetings, the dependencies, the noise—it's all still there.
And that's the tragedy: the first pod could have shown everyone what better really feels like. Instead, it risks teaching the opposite lesson—that pods are just another idea that doesn't work.
The leadership belief is real, the intent is strong, but without protecting the integrity of the experiment, the promise will vanish before anyone gets to see it fulfilled. This is how adaptive systems defend themselves—not with defiance, but with polite, well-intentioned erosion.
