Most leadership models still treat organizations as machines.

The metaphor runs so deep that most leaders do not know they are using it. Problems get categorized, responses get slotted to match, and the logic feels clean. It produces leaders who believe they can diagnose a problem, prescribe a solution, and predict the result — and organizations that prize control over adaptation, efficiency over resilience, and predictability over learning.

In stable conditions this works passably well. In turbulent, uncertain, novel and ambiguous conditions it does not just underperform. It shrinks the range of adaptive possibilities available to the system and locks the organization not into a zone of adaptability but into one of mediocrity.

How much less stable those conditions have become is worth stating as a number. The mean tenure of a company in the S&P 500 was 25 years in 1980. By 2012 it was 18. At the start of 2026 it is 15.

A company’s expected life on the index has contracted by two‑fifths inside a single working career, and the churn is accelerating rather than settling. That is the environment every leadership model now has to survive — and it is emphatically not the one the machine metaphor was built for. An organization optimized for predictability is optimizing for a world that is leaving.

Mean tenure figures: Torsten Sløk, Apollo, January 2026.

The uncomfortable part is that the models which adopted the language of adaptability — adaptive leadership, transformational leadership, agile — largely kept the machine’s bones. They still place the leader as prime mover: the one who gets on the balcony, regulates the distress, drives the change. These are exhortations dressed in complexity language but operating on linear assumptions. The language is new; the operating system is unchanged.

You cannot install culture the way you install software.

Leadership, culture and resilience are not things a leader creates. They are things that emerge, from countless interactions among people — the strength of their relationships, the roles they play, the rules they play by, and how they respond to signal, noise and disruption.

Call this complexity‑informed leadership: the discipline of treating the organization as the kind of system it actually is, rather than the machine it is usually managed as. Its most concrete expression is dissent‑friendliness — treating dissent the way a healthy ecosystem treats variation, not as noise to manage out but as the raw material resilience is built from.

The frame

Resilience is the strategy. Not performance.

Performance matters enormously. But performance without resilience is brittle, and an organization optimized for performance in stable conditions will shatter when those conditions change — which, now, they always do.

Performance is what the system produces on a good day. Resilience is what determines whether the system has good days at all, and how it recovers when it doesn’t.

Resilience is not a program, a training day, or a box to be ticked. It is a trajectory. The system is always either building it or eroding it, and the useful question is never are we resilient? but are we becoming more resilient, or less?

It has two faces, and both have to be cultivated. Robustness is the capacity to absorb a shock without breaking — the loss of a key person, a market turn, a failure in the field — and keep functioning. Regeneration is the capacity to recover, reconfigure, and come out of the disruption with capabilities the organization did not have before. Your immune system runs both: the innate response that resists immediately and generically, and the adaptive response that takes longer, learns, and remembers.

A team that can only resist will eventually meet a shock that overwhelms it. A team that can only adapt, with no stable core to anchor its identity, drifts into incoherence.

Treating resilience as the strategy changes the question every other decision gets evaluated against. Not what is the most efficient way to hit this quarter’s targets— but what conditions will let this system sustain performance, adapt to disruption, and grow stronger over time?

Held over time, that discipline shows up as specific things: better alignment between strategy, operating model, and behavior; fewer transformations that produce activity without durable change; more effective integration across functions or after an acquisition; less dependence on informal power structures or key individuals; and greater organizational capacity to absorb whatever change comes next.

The method

Don’t design the system. Design the conditions under which the system can design itself.

In a complex adaptive system you cannot script every move. You cannot know what challenges will arise, what combinations of conditions will appear, or what solutions the system will need. What you can do is shape the space in which resilient behavior becomes more likely — the conditions, the constraints, the feedback.

The leader here is closer to a gardener than an architect. Gardeners do not command plants to grow. They tend the soil, regulate water and light, remove the parasites, and trust the biology.

Three things a leader actually sets:

• Initial conditions. How a system begins profoundly shapes how it evolves. The simple rules established on day one, who is in the room, what the first meeting rewards and what it lets pass — these function like genetic code. And in human systems, unlike physical ones, we can start over. Initial conditions are revisable. That is the entire premise of doing this work at all.
• Boundaries. Not restrictions — focusing mechanisms. A pot shapes water; without the pot the water has no form. Without boundaries, energy dissipates and creativity diffuses. Time, resources, ethical and operational standards, and the non‑negotiables you will defend under pressure: these define the playing field without dictating the play.
• Shared purpose. A compass, not a map. A map assumes you already know the terrain. A compass gives direction in terrain nobody has seen, so sub‑teams can pick their own route up the mountain while staying oriented on the same summit. It is also what stops feedback degrading into personal preference — with a clear why, feedback can be anchored in is this aligned with our purpose?

This is not laissez‑faire leadership, and it is not standing back while the team figures it out. It is intensely involved. The involvement is in setting conditions, adjusting them, watching the signals and making calibrated interventions — not in prescribing solutions.

The leader’s actual instrument

Leadership is signal modulation in a living system.

Along the relationships in an organization, three things travel: information, energy and resources. Formal leaders sit at high‑traffic nodes, so their signals propagate further, faster, and with more amplification than anyone else’s — not because they matter more as people, but because of where they sit in the network.

And the signal is not what they say. Behaviors are the dominant signal. Words matter, but words without behaviors to back them up are noise — and where words and behaviors diverge, that is not inconsistency, it is hypocrisy. The system reads hypocrisy with exquisite precision. You can declare that you value feedback in every meeting; if the first person to deliver hard feedback is punished, the system learns instantly what is actually valued.

What this asks of a leader is coherence: that what they think, feel and do point the same direction. A tuning fork does not force other strings to vibrate. It emits a steady tone, and anything tuned to that frequency begins to resonate. Not through coercion — through coherence.

Coherence is not conformity. Coherence without variety is rigidity; variety without coherence is chaos.

Which is why leadership in this model is distributed rather than positional. Not as an egalitarian sentiment — as a structural necessity. When the unexpected arrives, the organization that can surface leadership from anywhere is the one that reconfigures in time. Distributed leadership builds redundancy: if one node goes down, others carry the load.

Where this meets the diagnosis

When you cannot know whether you are right, make being wrong cheap — and make it visible.

The diagnosis tells you where to act. It does not tell you how to shape the act itself when you cannot know in advance whether it will work. Two disciplines govern that, and they are partners.

This is the point where the two halves of the work turn out to be one discipline. A Soundings finding is not admissible unless it states what would prove it wrong. An intervention made under uncertainty is held to the same standard: the intervention‑side echo of a discipline the diagnostic work holds on its own side.

One caution carries all the way through. A convex, self‑invalidating move is still only as good as the system’s capacity to absorb it. A system already saturated cannot metabolize even a well‑formed intervention, and the move waits on the capacity to receive it. That is the same finding the diagnosis produces when it concludes no intervention is yet admissible — read from the leadership side rather than the diagnostic one.

How the diagnosis works

What else is in the model

The full framework is a worked body of material, developed over years and used in engagements rather than published. What sits behind the four ideas above:

• Interaction space The terrain where leadership, culture and resilience emerge or fail to — mapped as five elements: role players, relationships, roles, rules and response mechanisms. Shaping it is the leader’s most powerful dial. The same five are what the diagnosis uses to name a mechanism.
• Emergent engineering The method by which conditions get set and tuned rather than outcomes scripted, and the discipline that keeps it from collapsing into a checklist.
• Cultural dials Four everyday channels through which structure becomes lived experience: what gets rewarded, what gets ritualized, what level of engagement is genuinely invited, and what your behavior signals.
• The fractal pattern Why the same resilient architecture repeats from the individual to the team to the organization to its environment — and why higher levels are not permitted to break the rules of lower ones. Organizational resilience does not get bought with team dysfunction, and team resilience does not get bought with individual depletion.
• The attunement checklist Seven system vitals with a compass question attached to each, so a leader can read whether the system is trending toward resilience or fragility in real time — distinct from, and much lighter than, a full diagnostic sweep.

The honest reckoning

Why this fails

If this were as straightforwardly powerful as the pages above suggest, it would be everywhere. It is not, and the reasons are structural rather than a matter of anyone being stupid or lazy. Anyone taking this on should know them in advance.

• The gravity of linearity. The brain is a prediction machine, and under pressure it simplifies: categorize, make a checklist, reduce the system to parts. Leaders take dials and treat them as sequential steps, take a diagnostic and treat it as a scorecard. Each move kills the thing it tries to capture. Flatten the interdependence into a linear checklist and you have a corpse dressed in complexity language.
• Mistaking the map for the territory. The dials and the five elements are models — ways of seeing something that is in truth seamless. Treat them as independent boxes you can adjust one at a time and you lose the interdependence that made them worth having.
• The energy problem. This work is information‑ and energy‑intensive, and it cannot be delegated to a committee, outsourced, or accomplished in a weekend retreat. The leaders who most need it usually have the least bandwidth for it. The answer is not to try harder; it is to build the structures — distributed leadership, rhythms, working feedback — that sustain the work, because willpower alone will not carry it.
• The institutional immune response. Introduced into a linear organization, this looks like a foreign body. It challenges assumptions about control, redistributes power, values dissent and tolerates failure — and each of those triggers resistance. That response is not irrational; people have watched initiatives come and go. Which is why it cannot be imposed by mandate: introducing it is itself an exercise in it. Start small, with a willing team, and let the pattern spread by results and story.
• The humility gap. This requires the practiced acknowledgment that you cannot outthink a system — and leadership selection overwhelmingly favors confidence, decisiveness and the appearance of control. The leader who says I don’t know, but I trust the system to find the answer is rarely the one who gets promoted. But they are the ones who nurture resilience.
• The timescale mismatch. Trust, cultural health and adaptive capacity take months or years to show. Quarterly reviews and three‑to‑five‑year tenures demand results now. So leaders reach for the visible short‑term move — and each reorganization disrupts the relational networks that were slowly strengthening. Partial defense: measure system health as a leading indicator, and be explicit about the timescale up front.

None of these are reasons not to do the work. They are the reasons it has to be entered deliberately, with the failure modes named before the first move rather than discovered at month nine.

How the work runs

Engagements are taken with a small number of leadership teams at a time, because the work is iterative, specific to your system, and grounded in honest conversation rather than a delivery schedule.