Why I Do This Work: The Five Mechanisms

Ask someone in change work why they do it and the answer usually sounds like some version of "I want to help people." Fair enough. I love helping people and organizations too. But that answer is unfalsifiable, which makes it a weak response to a fair question.

Five specific things change inside an organization that actually gets better at change. None of them show up on the slide deck that goes to the executive board or even the steering committee.

A mechanism, in biology, is the pathway by which something actually happens, rather than a description of what happened. A forest floor stays damp through a dry summer because canopy shade slows evaporation, roots hold the soil open, and leaf litter keeps the ground cool. Nobody operates any of that. It runs on its own, continuously, which is the difference between a system that adapts and one that waits for someone to come and fix it. These five mechanisms work the same way.

Let's address three common objections first, because they are all reasonable and my argument is stronger after I address them. The first is that a consultant arguing organizations should need fewer consultants is either performing modesty or running a sales device. The second is that external expertise is often precisely the right choice, and an organization facing a once-in-a-decade ERP migration has no business building a permanent internal capability for it. The third argument comes from the evidence. Qualtrics surveyed 33,831 employees across 24 countries last fall and found that people who went through substantial organizational change reported higher engagement, not lower. New technology adoption tracked a ten-point increase. Layoffs tracked a seven-point decrease. Change volume is not the harm.

I would not dispute any of that. My argument is more narrow. Buying the same expertise repeatedly, for the same category of problem, is waste. What follows is what changes once the second purchase stops being necessary.

Mechanism One. The money stops being spent externally

The European Court of Auditors reviewed twenty external consulting contracts across the Commission's four highest-spending Directorates-General and reported that "none of the contracts we reviewed included a specific requirement for skills transfer from the consultants to the Commission staff." The Commission's own confirmation, quoted in the same report, is more blunt. Service contracts "rarely consider the transfer of knowledge from consultants."

That data came from inside 8,009 contracts worth €2.7 billion over three years, in an institution with more procurement discipline than most companies will ever have.

The UK's National Audit Office found the same problem and wrote it up as a formal lesson: do not become dependent on consultants, repeatedly using external consultants for the same tasks. Its explanation of why transfer fails is procedural rather than moral. Officials "often wait until the end of an engagement to talk about learning and knowledge sharing. At that point, consultants and civil servants are often focused on the next problem, and learning activities often get skipped or shortened."

What that freed money then does is something I see as intent rather than evidence. There’s no public benchmark for mid-market consultant spend being redirected into salaries and bonuses, and inventing one would be worse than acknowledging the discrepancy. The arithmetic is straightforward, though: a budget line that stops recurring becomes available for something else.

One practical difference worth knowing if you operate across both. France and Australia require knowledge-transfer clauses in consulting engagements, and France has required since January 2022 that civil servants be embedded in consultant teams wherever possible. The European Commission and UK central government require neither. Capability transfer is a contracting decision before it is a cultural one.

Mechanism Two. You move before you’re in crisis

Gartner surveyed 313 senior leaders and found that organizations which continuously adapt change plans based on employee responses are four times more likely to achieve change success.

The alternative is well documented. PwC surveyed 1,812 business leaders across 42 countries. Nine in ten had been through multiple major disruptions, averaging 3.5 in two years, and 76% called the impact medium or high. A full 70% were confident in their ability to recover. Most lacked the foundational resilience to do it. Confidence and capability had come apart and nobody on the inside had noticed.

Assessment and Sensing is the first of the six dimensions in the Adaptive Capability Ecosystem (ACE), my proprietary approach for building continuous adaptive capability, and it is first for a reason. The roots of that forest register a dry spell before the canopy shows any signs of it, and the tree starts conserving water while there is still water to conserve. An organization that cannot see what is coming has no way to prepare for it. It can only react in crisis mode, and by then it is paying for the delay as well as the problem.

Mechanism Three. The pace becomes survivable

Employee willingness to support enterprise change fell from 74% in 2016 to under 40% in 2022, while the average employee absorbed ten planned enterprise changes a year, up from two.

The reflex diagnosis is change fatigue and the reflex cure is better communication. Both are usually wrong. Gallagher surveyed 1,300 communications and HR professionals across 40 countries and found 61% have no formal approach to change communication at all, while simultaneously ranking change management their most valued communications skill. That is not a fatigue problem. It is a governance one.

This is where the Qualtrics finding earns a second mention. Employees are not worn down by change. They are worn down by change that goes nowhere.

They have seen this before. The Agile transformations of the last decades are the clearest case: new ceremonies, new job titles, a wall of sticky notes, and the same inefficient approval chain sitting underneath all of it. People learned to stand up for fifteen minutes every morning and report into exactly the same hierarchy they had before. Nothing about how decisions got made had changed. I even wrote a book about it.

AI is heading the same way, and for the same reason. It is being run as a technology project, which makes the tool the deliverable. A transformation is not a tool. It is a mechanism for changing how the business actually operates, and when the operating model comes out the other side unchanged, the organization has bought software and called it a transformation. Employees can tell the difference. They are usually the first to see it and the last to be asked.

Mechanism Four. The work actually changes

Bain studied 426 leaders of large-scale change programs. Only 12% achieved their full targets. Another 20% failed completely. The remaining 68% achieved more than half their goals and stopped short of their ambitions.

The 68% is the number that matters. The modal transformation is not a failure. It is paid for, partially built, and not delivering. Delivered and changed are not the same thing. That distinction is where most of the 68% lives.

An ERP program that goes live has been delivered. Whether it finished is a different question, and the answer sits in what happened to the work. Are processes more streamlined than they were? Do decisions get made with clarity, by people who can actually see what they need to see? Are workflows deliberate rather than inherited? Has waste come down? A system can be live, stable, fully adopted, and sitting on top of the same tangled process it was bought to fix. The same invoice still crosses the same desks. It just crosses them in a new interface.

Agile is the same test, one layer up from the ceremonies. Finished is not teams running stand-ups and shipping features a little faster. Finished is the business deciding what to build based on what customers actually want, budgets that move with the work instead of being locked twelve months ahead, teams and departments that can act without waiting for permission, and enough psychological safety that someone can say a thing is not working while it is still cheap to change. If none of that moved, the transformation delivered. Nothing changed.

McKinsey put a figure on what that costs. Successful transformations capture 67% of the financial benefit available to them, everyone else averages 37%, and almost a quarter of the loss happens during target-setting, before the program has started.

For a European reader the sharpest numbers come from Horváth, which surveyed 200 companies with at least 200 employees and €200 million in sales across DACH, Northern and Eastern Europe and the US. Transformations ran 30% longer than planned. More than 60% deviated on budget, schedule and result quality.

Hundreds of thousands, and often millions, spent on something that arrived late and did most of what it promised.

Mechanism Five. People see what they built

Bordia and colleagues tracked this properly, over two years, in peer-reviewed work. A poor change management history shaped what employees believed about change. Those experiences lowered trust, satisfaction and openness, raised cynicism, and then predicted actual turnover across the following two years. Actual employee turnover. Not just dissatisfaction.

Only 17% of organizations reported high maturity at benefits realization in PMI's 2016 benchmark, the last time it measured this. Where benefits are identified before a project starts, 74% meet their goals and business intent. So the practice that most reliably makes a transformation pay off is the one five out of six organizations have not built.

For European organizations the stakes sit higher, not lower. Gallup has found European workers the least engaged regional workforce five years running, most recently at 13%. An abandoned initiative costs more against that baseline, not less.

Bordia's study measured the damage rather than the repair, but the mechanism it found runs through what employees believe about change, and beliefs move in both directions.

What moves them is unglamorous and mostly about who was in the room. The people who do the work are involved while the decisions are still open, not consulted once the design is locked. What leadership promised at the start is what shows up at the end, or the change to the promise gets explained by someone willing to own it. Leaders behave as stewards of the thing rather than sponsors of it, which mostly means continuing to nurture the conditions needed for success.

An organization that manages that once has a workforce with evidence. The next initiative starts somewhere different, because people have been through one that went where it said it would go. That is the compounding part, and it is why adaptability is a capability rather than a project. Every transformation people live through well makes the next one cheaper.

What the work is for

Five mechanisms, and not one of them is "helping people." Each is a thing that either happens or does not, and each can be checked by someone who wants to check it.

That is my whole answer. I still love helping but I do it with intention, measurement, and results.

If any of this describes something recognizable in your own organization, the Organizational Adaptability Pulse Check is twelve questions and takes about five minutes. It gives a starting read on where your organization actually stands. For leaders who already know roughly where they stand and would rather talk it through, a conversation is the more useful next step.