Organisational Drift

In 1996, a sociologist named Diane Vaughan published a book about the Challenger space shuttle disaster that changed how we think about organisational failure. She'd spent nine years studying NASA's culture before that January morning in 1986 when seven astronauts lost their lives. What she expected to find was misconduct - managers cutting corners, engineers being silenced, someone somewhere making a reckless call.

What she found instead was far more unsettling.

There were no villains. No conspiracy. No reckless decisions. What Vaughan uncovered was a slow cultural process in which concerns about a known design flaw - the O-ring seals on the solid rocket boosters - were gradually absorbed into normal operations. Each time a shuttle launched with the flaw and nothing went wrong, the boundary of what counted as "acceptable" shifted slightly. Not through any deliberate choice, but through the quiet logic of repetition.

Her term for this was the normalisation of deviance: the gradual process through which unacceptable practice or standards become acceptable, simply because they've been repeated without catastrophic results. She described a "long incubation period" in which each small deviation made the next one easier to accept. Every time a rule was bent and nothing bad happened, it became a little harder to argue that the rule still mattered.

You probably don't work at NASA. But the pattern Vaughan identified isn't about rockets. It's about something structural in the way organisations operate - and it's happening, right now, in organisations of every size and sector. Including, quite possibly, yours.

What drift looks like in everyday organisational life

There's a version of this pattern that shows up across the organisations we work with, and it rarely looks dramatic. It looks like this:

A leadership team that started out genuinely collaborative gradually becomes a group of people managing their own areas and occasionally updating each other. A decision-making process picks up an extra approval stage every year until nobody can remember what speed felt like. A relationship with the people the organisation exists to serve becomes progressively more distant, more mediated, more managed - without anyone choosing that direction. These are the patterns that show up in culture change work - not because people failed, but because the system shifted around them.

In systems thinking, this pattern is called drifting goals or eroding goals - a dynamic first mapped by Peter Senge in the early 1990s. The mechanism is elegant and a little uncomfortable. When there's a gap between where an organisation is and where it wants to be, two options exist: take corrective action to close the gap (hard, slow, uncertain) or quietly lower the standard (easy, immediate, tension-relieving). Under pressure, the standard almost always gives way.

Each individual lowering makes sense at the time. The rationalisation is familiar: we'll get back to the original standard once things settle down. But things don't settle down. The new, lower standard becomes the baseline from which the next gap is measured, and the cycle repeats.

There's an almost comic illustration of this from the systems thinking literature. A food company committed to producing a high-quality product began a series of cost-cutting measures. Line speeds were increased, cooking methods changed, storage adjusted. Over several years, sales declined. Managers attributed the decline to changing consumer tastes. It took formal consumer research to reveal what had been happening in plain sight: the product had quietly become worse. Nobody had made a single decision to lower quality. It happened through a sequence of efficiency improvements, each one perfectly reasonable on its own.

Why drift always moves in one direction

Here's the structural question that makes drift so persistent: why does it always seem to erode in the same direction? Why do standards slip rather than spontaneously rise?

A safety researcher named Jens Rasmussen offered an answer in 1997 that applies far beyond the safety domain. He described organisations as operating within a space defined by competing pressures. On one side, there's the pressure for economic efficiency - do more with less, move faster, cut costs. On another, there's the pressure of workload - what people can reasonably sustain. And on a third side, there are the boundaries of quality, culture, values, and service standards.

The critical asymmetry is this: efforts to improve efficiency tend to produce gains that are immediate, tangible, and visible. But the erosion of culture, quality, or values produces feedback that is delayed, ambiguous, and easy to miss. The visible always wins against the invisible. Not because anyone chooses efficiency over values, but because efficiency makes its case loudly and in real time, while the erosion of something like organisational culture happens so quietly that it doesn't register as a signal at all.

The sociologist Karl Weick captured this beautifully: reliability is invisible, so there is nothing to pay attention to. You could substitute "culture" or "service quality" or "strategic coherence" for "reliability" and the observation would hold just as well. The things that matter most in organisational life are often the things that generate the least visible feedback - until they're gone.

This is why drift is so hard to detect from inside the system. The efficiency gain from this quarter's restructure is obvious. The slight erosion of collaborative culture that accompanied it won't show up for eighteen months - and when it does, it'll present as a "communication problem" or an "engagement issue" rather than as a consequence of that earlier decision.

How drift compounds across an organisation

In 2000, a military researcher named Scott Snook published an analysis of one of the most baffling incidents in modern military history. In 1994, two US Air Force fighter jets accidentally shot down two US Army helicopters over Iraq, killing twenty-six peacekeepers. The investigation that followed had virtually unlimited resources and took almost two years. No one was found culpable. No equipment had malfunctioned. No rules had been deliberately broken.

What Snook found was something he called practical drift: the slow, steady uncoupling of practice from written procedure. Each unit involved - the fighter pilots, the helicopter crews, the air traffic controllers - had adapted their routines in small, locally sensible ways. None of these adaptations was dangerous on its own. But they had compounded across the system in ways that nobody could see from any single vantage point.

His conclusion is one that should give every leader pause: this accident happened because - or perhaps in spite of - everyone behaving exactly the way we would expect them to behave.

This is the dimension of drift that's hardest to sit with. It's not just that standards erode gradually. It's that they erode differently in different parts of the organisation, and nobody has the complete picture. The finance team's version of "how we do things" has drifted in one direction. The frontline's version has drifted in another. The leadership team's assumptions about what's happening on the ground may bear increasingly little resemblance to what's happening on the ground. And because everyone is operating from their own locally coherent version of reality, the divergence doesn't surface until something goes wrong. This is what organisational silos are telling you - not that people are being territorial, but that different parts of the system have drifted in different directions without anyone noticing.

This is the compounding effect of drift. It's not one thing moving slowly. It's everything moving slowly, in slightly different directions, with no mechanism for anyone to notice the accumulation.

The organisation as an intelligence of its own

There's a more recent thread of thinking that takes this further. In 2024, the economist Dan Davies published an argument that organisations should be understood as systems capable of producing decisions that are distinct from the intentions of any of their members. He describes what he calls accountability sinks - structures within organisations that absorb or obscure the consequences of decisions, severing the link between what happens and who's responsible for it.

This connects to drift in a way that's worth pausing on. When nobody is specifically accountable for maintaining a particular standard - when it's everyone's job in general and nobody's job in particular - that standard becomes vulnerable. Not to neglect, exactly, but to the quiet gravity of everything else that's competing for attention. The accountability sink doesn't cause the drift. But it removes the feedback mechanism that would make the drift visible.

Davies puts his finger on something that many leaders recognise: organisations routinely produce outcomes that everyone involved claims not to want. Nobody decided to let the culture erode. Nobody chose to let service quality slip. Nobody planned for strategy to become disconnected from operations. And yet here we are. The system produced these outcomes through the accumulated weight of reasonable decisions, each one defensible on its own terms.

What makes some organisations more susceptible

Not all organisations drift at the same rate, and the research suggests some conditions accelerate it.

Organisations under sustained pressure to deliver - merger integration, regulatory compliance, rapid growth - are particularly vulnerable. The urgency of immediate demands makes it harder to tend to the slower, quieter things that keep the organisation coherent. Sidney Dekker, a safety researcher who built on the work of Vaughan and Rasmussen, made a provocative observation: organisations drift into failure precisely because they are doing well. Success breeds confidence. Confidence reduces the perceived need for vigilance. And reduced vigilance is the condition in which drift thrives.

Organisations with weak feedback loops are also more exposed. If frontline intelligence doesn't reach the people making strategic decisions - if the signals that something is shifting get filtered, softened, or lost on the way up - drift can continue unchecked for years. The organisation's formal metrics may all be green while the lived experience of working there, or being served by it, has quietly deteriorated.

And organisations going through leadership transitions are at particular risk. A new leader inherits an organisation as it is, not as it was designed to be. Without an anchor in the original intent, the drifted state becomes the baseline. The erosion gets locked in as "how things work here" rather than being recognised as a departure from something that came before.

The question that changes

There's something generous in this body of research. Across Vaughan, Rasmussen, Snook, Senge, Dekker, Davies - across decades of study in fields as different as sociology, safety science, military analysis, and economics - the same finding recurs: drift contains no villains. People aren't failing. The system is producing outcomes that nobody intended through the accumulated weight of individually reasonable decisions.

That's not a comfortable finding, but it is a liberating one. Because if drift isn't caused by bad people, it can't be fixed by finding better ones. The response to drift isn't to work harder, hire smarter, or hold people more tightly accountable. It's to build the conditions that make drift visible before it compounds - to create the feedback loops, the honest conversations, the regular moments of stepping back and asking: is this still what we intended?

The question at the heart of drift isn't "who let this happen?" It's something quieter, more honest, and more useful: what did we set out to be, and how far have we travelled from it?

That second question is where some of the most important work in organisations begins. Not because the answer reveals failure. But because it reveals the distance between intention and reality - and in that distance, there's always something worth understanding.