Organizations Are Human Systems, Not Machines

Organizations are usually managed as though they were machines. A problem appears, so someone redesigns a process, introduces a policy, schedules a meeting, builds a dashboard, implements new software, clarifies roles, updates the metrics.

Sometimes it helps. Often it doesn't — and not because the process was badly designed. Because an organization is not a mechanical system in the first place.

A machine behaves predictably. Replace a worn part and performance changes in a knowable way. Human systems don't work like that. The same policy produces different outcomes in different teams. The same leader inspires one group and grates on another. The reason is simple and consequential: people don't just execute processes. They interpret them.

This is the core of what organizational theorists mean by sensemaking. Organizations function as interpretive systems — people continuously assign meaning to what happens around them, and then act on the meaning rather than on the event (Weick, 1995). A reorganization isn't experienced as a new box on a chart. It's experienced as what this says about my standing, my future, my safety. The chart changed one thing. The interpretation changed everything downstream.

The two systems

Every organization runs two operating systems at once.

The first is visible: org charts, reporting lines, budgets, technology, workflows, policies, performance metrics. It's the system most management attention goes to, because it's the one you can see and edit directly.

The second is largely invisible: trust, assumptions, relationships, informal influence, psychological safety, reputation, identity, shared history. No one designs this system on purpose, and it governs how the first one actually behaves.

Trying to run an organization through the visible system alone is like trying to understand traffic by studying the roads while pretending the drivers don't exist. The roads set constraints. The drivers determine what actually happens on them. This is why two companies with nearly identical structures, strategies, and tools produce dramatically different results. The structure didn't create the difference. The people interpreting and inhabiting it did.

Communication isn't an information problem

Watch what this explains that the machine model can't.

Organizations routinely diagnose their communication failures as information failures — people didn't get the memo, the update wasn't clear, the tool wasn't used. So they add more channels, more updates, more dashboards. And the problem persists, because most communication breakdowns have very little to do with the availability of information.

People stay quiet when they have information leaders need, because they've calculated the consequence of speaking. Leaders avoid the direct conversation because conflict is uncomfortable. The information exists and is perfectly clear. Whether it moves is determined by the invisible system — by whether it feels safe to say the true thing. When enough people withhold what they actually think, the organization operates on a sanitized picture of its own reality, and no communication platform fixes that, because the platform was never the constraint.

Why change is really resisted

The same misreading drives most failed change efforts.

When people resist a change, organizations tend to assume the message wasn't clear enough or the training wasn't good enough — an information problem again. Usually it's something else. People are answering a question the rollout deck never addressed: what does this mean for me?

That question is not vague anxiety. It's specific, and the specifics are now fairly well mapped. The brain treats certain social stakes as survival-level concerns — among them status, certainty, autonomy, and relatedness — and a threat to any of them triggers a genuine threat response, not a rational objection to be argued away (Rock, 2008). Read resistance through that lens and it resolves into concrete questions. Will I lose influence? Will I look incompetent at something I used to be good at? Will I have less control over my own work? Will this cost me the relationships I rely on?

A leader who hears "resistance to change" reaches for more communication. A leader who hears "status, certainty, autonomy, and relatedness are under threat" does something different — they address the actual stakes. The change hasn't changed. What the person needed to know about it was never in the training.

Culture is what gets reinforced, not what gets stated

The invisible system has an output, and we call it culture.

Culture is not produced by mission statements or values painted on the wall. It emerges from repeated interactions — from what people actually experience, over and over, about how this place works. And people are shrewd readers of the difference between stated and enacted values. They don't follow what an organization says it prizes. They follow what it consistently rewards, tolerates, and punishes. A company can state that it values candor and reward agreeableness, and everyone will quietly learn the real rule within weeks.

This is why culture change by announcement never works. The announcement lives in the visible system. The culture lives in the invisible one, and it updates only when the pattern of real, repeated experience updates.

What this changes

None of this makes structure, process, or metrics unimportant. The visible system is real and necessary. The error is believing it's the whole machine — that if you get the boxes and the incentives right, the human behavior will follow. It won't, because the behavior was never running on the boxes. It was running on interpretation, safety, identity, and trust.

Organizations don't think. People do. Organizations don't build trust, or resist change, or withhold the truth in a meeting. People do — for reasons that make complete sense once you can see the second system.

So the most useful question a leader can ask is rarely what process would fix this? It's usually what are people experiencing here that makes this behavior the sensible response? Answer that, and the fix is often obvious. Skip it, and you will redesign the visible system forever while the invisible one quietly does whatever it was always going to do.

Organizations are not machines to be engineered. They are human systems to be understood.

References

Rock, D. (2008). SCARF: A brain-based model for collaborating with and influencing others. NeuroLeadership Journal, 1, 1–9.

Weick, K. E. (1995). Sensemaking in organizations. Sage.