Self-organization is probably the most misunderstood concept in modern organizations

If there's one topic that almost inevitably comes up whenever I talk to executives about modern organizations, agile ways of working, the adaptive organization or similar, it's self-organization.

At some point, the conversation shifts away from Scrum, Kanban or product development and becomes much more fundamental. People talk about empowering teams, flattening hierarchies and pushing decisions closer to where the work happens.

The underlying assumption is almost always the same: if we remove enough bureaucracy and give people enough autonomy, better decisions and better outcomes will naturally follow.

For a long time, I believed that too. Today, I'm less certain and more reflective about that assumption.The reason isn't theoretical. It's because I've watched organizations pursue exactly that goal, often with the best intentions, only to find themselves six or twelve months later wondering why they had become slower rather than faster.

One engagement has stayed with me.

The company had launched what they proudly called an "empowered organization." Managers had been encouraged to step back, approval processes had largely disappeared, and product teams were told that they now owned their products end-to-end.

It sounded exactly like the kind of transformation many organizations aspire to. When I returned a few months later, I joined one of their portfolio meetings. It didn't take long before a familiar pattern emerged.

A discussion about an important customer feature quickly turned into a search for ownership.

• Engineering believed Product Management had the final call.
• Product Management assumed Architecture needed to decide.
• Marketing had already announced a release date because nobody had objected.
• Someone eventually asked whether Team A or Team B was responsible. Nobody seemed entirely certain.

What struck me wasn't the confusion itself. Large organizations are complex and uncertainty is part of everyday business. What surprised me was that every person in the room genuinely wanted to do the right thing. Nobody was hiding. Nobody was avoiding accountability. Yet the organization had become noticeably less effective.

See also: What We Got Wrong About Self-Management

As I reflected on that meeting afterwards, it became clear that the company hadn't failed because it had given people too much autonomy.

It had failed because it had removed a decision-making system without replacing it with another one.

That experience fundamentally changed how I think about self-organization.Today I see it less as something you implement and more as something that emerges when the surrounding conditions are right.

Structure is not the system

Unfortunately, that's not how many organizations approach it. Over the years I've seen countless companies look at successful organizations like Spotify or Zappos, or experiment with concepts borrowed from Holacracy or Sociocracy.

See also:

• Exploring The Spectrum Of Self-Management: From Holacracy To Co-Elevating Teams
• Why is Zappos’ culture of holacracy catching on?

The visible parts of those models are naturally attractive. Squads, circles, chapters and distributed leadership are easy to recognize, easy to explain and relatively easy to copy.

The invisible parts like shared context, decision discipline, mutual trust and organizational capability are much harder.

I remember speaking with an executive who proudly explained how they had replaced traditional management structures with circles inspired by Sociocracy. Curious to hear about the results, I asked him what had happened after the initial rollout.

He laughed.

"After about four months," he said, "we realized we'd created managers again. We just stopped calling them managers."

At first, we both found it amusing. But the more I thought about it afterwards, the more profound that observation became. People naturally look for orientation. Whenever a system no longer provides clarity about how decisions are made, organizations have an incredible ability to recreate that clarity informally.

Leadership doesn't disappear. It simply becomes unofficial, less transparent and often less accountable than before.

The problem wasn't Sociocracy.

The problem was believing you could separate a handful of attractive practices from the broader operating system that made those practices successful in the first place.

Perhaps the strongest counterexample I've encountered comes from startups. Founders rarely spend time discussing self-organization. I've never seen a founder gather everyone around a whiteboard to debate empowerment models or distributed authority. They're usually too busy building products, talking to customers, hiring people, and trying to survive another quarter. Yet many startups operate with remarkable autonomy.

I don't think it's because entrepreneurs possess some unique talent for self-organization. I think it's because they live in an environment where the conditions therefore already exist.

• They understand the customer because they speak to them every day.
• They understand the business because every commercial decision affects their runway.
• They understand the strategy because they created it.
• And when difficult decisions need to be made, they don't have the luxury of waiting for another committee.

Purpose, ownership, and context naturally come together.

Autonomy is a capability, not a permission

Large organizations often lose exactly those qualities as they grow. Work becomes increasingly specialized; ownership is distributed across departments and very few people still see the entire picture. Ironically, we often respond by increasing autonomy without increasing the capabilities required to use it well. That, in my experience, is where many Agile (or basically any other “ways of working”) transformations begin to struggle.

I once worked with an executive team that couldn't understand why their product teams kept escalating relatively small decisions.

"We've empowered them," one executive said during a workshop. "Why do they still keep asking for permission?"

When we spent time with the teams, the answer became obvious.

• Several (so-called) product owners admitted they had never really learned how to prioritize competing stakeholder demands.
• Technical leaders were highly capable engineers but had little exposure to commercial decision-making.
• Some team members openly admitted they didn't fully understand how the company's business model actually worked.

The escalation suddenly made perfect sense. It wasn't a lack of courage but a lack of confidence. We often assume self-organization is primarily about mindset. I increasingly believe it's about capability.

See also: The Buurtzorg Model: Self-Managing Teams Explained | Peerdom

Decision-making, prioritization, conflict resolution, commercial awareness and systems thinking aren't innate qualities. They're professional skills that need to be developed deliberately. Giving people greater authority without helping them develop those capabilities doesn't create ownership.

It simply exposes the gap that was already there.

Freedom needs a frame

The other lesson I've learned is that autonomy depends just as much on boundaries as it does on freedom. That may sound counterintuitive. Many people hear the word autonomy and immediately think fewer rules. Yet some of the fastest organizations I've worked with operated within remarkably clear constraints. One client completely redesigned the way they governed product development.

Instead of approving individual features, leadership agreed on a small number of non-negotiable principles:

• Customer outcomes
• Regulatory compliance
• Security standards
• Quarterly business priorities

Everything else became the responsibility of the teams. What surprised everyone wasn't that teams made more independent decisions. It was how much faster those decisions became. Nothing magical had happened. People simply spent less time wondering whether they were allowed to act.

I've like to think of this as “freedom within a frame”. The frame doesn't restrict autonomy but makes autonomy possible. Another realization came during a workshop with a utility company.

A manager challenged me directly.

"So are you saying we're not Agile because we still have approval gates?"

Instead of answering, I asked him what would happen if one of their engineers approved the wrong software update.

He smiled. "Millions of customers lose electricity."

That answer was all either of us needed. Not every environment benefits from the same degree of autonomy. A startup searching for product-market fit faces a very different reality from an emergency department, an airline operations center or a power grid operator.

Context matters far more than ideology.

See also: Medium and Holacracy: What We Can Learn from the Experiment | SI Labs

The best organizations don't try to maximize autonomy. They try to find the level of autonomy that best fits the work they're doing.

The goal was never self-organization

Looking back, I think we've spent years asking the wrong question. We've asked how organizations can become more self-organized. Perhaps we should have been asking how organizations become better at making decisions. Because that's really what we're trying to achieve.

Self-organization isn't the destination. It's a consequence.

When people share a clear purpose, understand the business they're in, possess the capabilities to make good decisions and operate within sensible boundaries, autonomy stops being a management initiative.

It simply becomes the way the organization works.

And perhaps that's the real lesson I've learned after two decades of organizational transformations:

Great organizations aren't built by removing leadership or structure. They're built by creating the conditions in which good decisions happen naturally, consistently and at every level of the organization.