Leading From Inside The System

The systems-aware leader’s case for stewardship over architecture

You can’t change a system you’re not part of.

You can’t change a system you’re not part of. So says Dr. Larry Krafft, Professor Emeritus at Temple University, one of the sharp minds who participated in The Ecosystem Project Learning Circle. When I first heard it, I heard it as a caution. You can’t parachute in. You can’t redesign from above. The system will absorb you, route around you, or reject you entirely. Outside expertise, however sophisticated, doesn’t move human ecosystems. It was a warning for consultants and strategists who mistake proximity for membership.

But lately I’ve been reading it the other way. If you can’t change a system you’re not part of, then being part of one means you can change it. That’s not a small thing. For leaders who have spent years watching their most careful efforts produce modest results, it’s a profound source of agency. The system is not immovable. You are not powerless. The question is whether you understand the power you actually have.

You need to think and act like a steward, not an architect

Think about the ecosystems that shape daily life in most communities: a waterfront district where public agencies, private developers, cultural organizations, and community groups all have a stake; a regional housing system with dozens of funders, builders, advocates, and government bodies pulling in different directions; an emerging AI ecosystem where startups, researchers, regulators, ethicists, and established institutions are all making consequential moves at once; a community foundation trying to shift the conditions that keep social problems entrenched. In every one of these, someone is nominally in charge. And in none of them does that person actually control what happens.

This is the structural reality of human ecosystems. They are decentralized. Their inhabitants and orchestrators act on their own intentions, pursue their own purposes, and extract their own value. What holds a system together is not hierarchy but the web of value exchanges among people, organizations, and all the transactions and interactions between them. That web creates a kind of order, a dynamic equilibrium, but it’s not an order anyone designed. It emerged from countless interactions over time.

This means that even a leader granted supreme authority over a complex ecosystem, a foundation, a municipal system, a regional social infrastructure, still cannot direct it the way you’d direct an organization. The variables are too many, the interactions too nonlinear, the emergent behaviours too unpredictable. The ecosystem continues to self-organize around its own logic, regardless of any organizational chart, policy, procedure, or program.

The pressure to act like an architect is understandable. To announce a strategy, restructure the parts, and design the intervention. It feels like leadership. It looks like leadership. But applied to a living ecosystem, it’s a tool built for a different kind of system entirely. The ecosystem will simply route around it.

And yet the pressure is real. The board wants a strategy. The funders want impact metrics. The public wants to see leadership. In a sustained moment of turbulence, and most senior leaders are in one right now, the pull toward visible, planful action is not weakness. It’s the entire culture of institutional leadership pointing in one direction.

The systems-aware leader’s dilemma is not a personal failure of nerve. It’s a structural tension between the accountability culture leaders operate inside and the ecosystemic reality they’re actually navigating.

To lead effectively from inside, you need navigational intelligence

Systems stewardship is not passive. It is not watchful patience dressed up in sophisticated language. It is the active, skilled work of helping a human ecosystem become more fit for its own purpose. And it operates on two levels simultaneously.

Some of what stewardship requires is planful: clear decisions, coordinated action, well-designed initiatives with measurable outcomes. Ecosystems contain components that respond to direct intervention. Not everything is irreducibly complex. Part of the steward’s skill is recognizing what can be planned and executing it well.

But ecosystems also contain dynamics that don’t yield to planning: emergent patterns, shifting relationships, trust that accumulates or erodes through a thousand small interactions. These require a different kind of engagement. Not a plan but a capacity to read what’s happening and recalibrate in real time. We call this navigational intelligence: the ability to read the invisible dynamics of a complex system, the shifting landscape, the diverse inhabitants, the underlying flows of value, and adjust your heading as you move. Not a map. Not a script. A capacity to navigate.

Most leaders over-index one way or the other. Some are rigidly planful, convinced that everything must cascade from a formally structured strategic plan, every goal made Specific, Measurable, Achievable, Relevant, and Time-bound. When conditions shift, as they always do, the plan no longer fits, but the leader doesn’t shift with it, and the gap between the plan and reality steadily widens. Others lean so far toward the adaptive that they shift with every new input, leaving their teams and boards uncertain about where the whole enterprise is actually headed. The effective steward holds both. Plans where planning works. Stays adaptive where the system requires it. Knows the difference, and is honest about which is which.

The accountability trap

There is a version of impact measurement that is genuinely useful. Understanding whether a system is moving toward health, whether value is flowing where it should, whether inhabitants are better supported, whether blockages are clearing, that’s not bureaucratic noise. It’s navigational intelligence in action. It tells you whether your heading is right.

But most accountability frameworks weren’t designed for ecosystems. They were designed for programs. They measure outputs, not dynamics. They reward legibility over complexity. They make it structurally easier to fund a discrete initiative than to support the ongoing, relational work of holding a system together.

Leaders who want to steward complex systems while operating inside accountability structures built for simpler ones are navigating a genuine tension. The answer isn’t to reject accountability. It’s to develop a richer language for what health in a human ecosystem actually looks like, and to bring funders, boards, and partners into that language over time.

That work is itself deeply relational. Systems don’t shift because one leader has a better map. They shift when many of the people inside them, across organizations and roles and sectors, can begin to see what’s actually happening and act from that shared understanding. Other leaders in adjacent organizations are making their own moves, responding to their own pressures, shifting their own corners of the ecosystem. When the shared picture is rich enough, those moves begin to reinforce each other rather than cancel out. Change in human systems is fundamentally a dialogic process. It happens in conversation, through relationships, across the ecosystem, over time.

Which means the steward’s most important work is often not deciding. It’s creating the conditions in which others can see more clearly and then act.

Agency, not control

The stewardship model is not a retreat from accountability. It is not an excuse for ambiguity or a rationale for inaction. It is a more accurate account of how human systems actually change, and therefore a more honest basis for leadership.

You can help the people inside a system see it more clearly. You can create conditions for better conversations. You can notice where value is flowing and where it isn’t, name what others sense but haven’t said, plan what can be planned and hold loosely what can’t. None of that is passive. All of it requires skill, patience, and a tolerance for the kind of leadership that doesn’t always look dramatic from the outside.

Dr. Krafft’s provocation cuts both ways. You can’t change a system you’re not part of. But if you are part of one, truly part of it, in relationship with its inhabitants, with a clear view of its dynamics, you have more capacity to shift it than any outside intervention could produce.”

The question isn’t whether you have agency. The question is whether you’re willing to use the kind you actually have.