The Safety of Change
Why Transformation Fails and How Sidney Dekker Can Help
Organisational transformation initiatives often flounder or fail because those impacted perceive the change as a threat to their safety in some way. How can we use the idea of safety to think about how we transform organisations?
When considering introducing change to an organisation, leaders often reach for traditional change management tools and tactics aligned with John Kotter’s 8-Step Model or similar frameworks. In a future post, I will describe my approach to change in large organisations (and it certainly isn’t an 8 Step process…) but first I need to lay the foundations by introducing some critical thinkers and ideas. Throughout this series I will mention links to thinkers that are relevant and will enhance them as the foundations are laid.
In this post, I briefly discuss Sidney Dekker.
Sidney Dekker is a giant in the field of safety science, best known for analysing plane crashes and medical errors. However, his work on ideas such as Local Rationality, and Drift provides a practical framework for thinking about organisational transformation.
The question that Dekker’s work brings to mind is: What if the reason for transformation failure isn’t execution, but a gap in our understanding of how complex human systems fail?
Here is how to apply Dekker’s “view from below” to save your transformation from drifting into failure…
1. Mind the Gap: Work-as-Imagined vs. Work-as-Done
In every organisation, there are two distinct realities. There is Work-as-Imagined; the formal procedures, the transformation roadmap, and the new org chart designed in the boardroom. Then there is Work-as-Done; the messy, adaptive, everyday reality of how people actually get things out the door.
Leaders often confuse the map with the territory. They assume that because a new process has been “rolled out,” it is being followed. When the transformation hits a snag, they enforce stricter compliance, or print more posters, or have more townhalls. Strict compliance with a rigid transformation plan can actually reduce performance. As Dekker notes, if people follow rules literally without their usual adaptive workarounds, the system often grinds to a halt.
2. Stop Blaming “Resisters”: The Principle of Local Rationality
When a team fails to adopt a new tool or reverts to legacy behaviours, it is easy to label them as “resistant to change.” Dekker calls this “The Bad Apple Theory”—the idea that the system is safe, and problems are caused by erratic individuals.
We investigate failure by asking, “Why didn’t they follow the plan?” This assumes the plan was fine and the people were the variable.
Dekker’s core principle is that people do not come to work to do a bad job. Their behaviour - even when it wrecks your timeline - made sense to them at the time, given their goals, focus of attention, and the tools available.
Instead of judging non-compliance, reconstruct the context. What pressures were they under? What information was missing? Did the new “efficient” process actually remove some process they see as critical?
We need to create the kind of psychological safety described by Gene Kim in The Unicorn Project, where safety is a prerequisite for honesty and improvement.
3. Detect the “Drift” Before the Crash
Transformations rarely fail overnight. They suffer from Drift into Failure. This is a gradual process where small, locally rational adaptations, as we described above, accumulate over time until the system crosses a boundary it didn’t know was there.
We look for big explosions or immediate rejection of the new strategy. We miss the subtle “normalisation of deviance”—where teams make tiny compromises to keep the new system running (e.g., skipping a daily stand-up, bypassing a security check), eventually migrating the organisation far from its intended design.
Drift is often driven by production pressure and resource constraints. If you cut budgets while demanding transformation, you are practically engineering drift.
This links to something we will be discussing later, Goldratt’s Theory of Constraints: Drift often occurs at the bottleneck (constraint) where the pressure to “exploit” the resource is highest.
4. Move from Safety-I to Safety-II
Traditional management focuses on making sure as few things as possible go wrong (Safety-I). Dekker and his colleagues propose Safety-II: ensuring as many things as possible go right. In my experience, this is quite an unusual perspective for transformations, especially those focused on constraints, like cost-cutting.
Transformation governance often consists of Red/Amber/Green status reports focused on defects, delays, and budget overruns. It defines success as the absence of problems.
Do not define the success of your transformation by the lack of complaints. Define it by the presence of adaptive capacity - the ability of the organisation to handle surprises and variations.
Don’t just investigate the failed pilot projects. Investigate the successful ones. How are they navigating the complexity? What “unsanctioned” adaptations are they using to make the new operating model work?.
Ron Westrum’s Generative Culture is very relevant here. His focus is on the mission and inquiry rather than bureaucracy and silence. We will be discussing him in future too.
5. Build a Just Culture (Restorative vs. Retributive)
This is one of my favourite ideas and is a very powerful operating principle. How you react when a migration fails or a deadline is missed defines the culture of your transformation. Dekker advocates for Restorative Justice (repairing trust and systems) over Retributive Justice (punishing individuals). Think about what this means for how we treat those who we think “just don’t get it”.
“Who messed this up?” “Who is accountable for this delay?” These questions trigger defence, causing people to hide the “weak signals” you desperately need to see.
If you replaced the “failing” team with another similarly trained team in the exact same situation, would they likely have done the same thing? If yes, you have a systemic transformation issue, not a people problem.
W. Edwards Deming, famously argued that 94% of troubles belong to the system (responsibility of management), and only 6% are special causes attributable to workers.
(An Organisational Prompt is something you can do now….)
The Organisational Prompt:
Create an initiative to identify and collect the weak signals that RESULT from how we treat failure or lack of progress in the transformation.
The idea here is simple. We tend to privilege the perspective of the plan over the perspective of the practitioner. We do this through enforcement. The weak signals are the conversations, the justifications and the little jokes that people tell in the wake of feedback about their performance in terms of the plan.
Further Reading
Sidney Dekker: Sidney Dekker’s Official Site - For deep dives into The Field Guide to Understanding ‘Human Error’ and ‘Drift into Failure’.
Gene Kim: IT Revolution - For The Phoenix Project and The Unicorn Project, which operationalise many of Dekker’s safety principles into DevOps.
Ron Westrum: A Typology of Organisational Cultures - The source of the Generative/Pathological culture model.
Dave Snowden: The Cynefin Company - For understanding complexity and the “Probe-Sense-Respond” approach that complements Dekker’s view of emergence.
John Kotter: Kotter International - For the traditional 8-Step Process for Leading Change.
Eliyahu Goldratt: TOC Institute - For the Theory of Constraints and understanding bottlenecks.
Disclaimer
I write about the industry and its approach in general. None of the opinions or examples in my articles necessarily relate to present or past employers. I draw on conversations with many practitioners and all views are my own.