Adopted, Never Adapted: How Governance Frameworks Ossify
Adoption is an event with a date and a launch; adaptation is a discipline with no end.
Executive Summary
Across a working life spent inside change programmes, one pattern recurs with almost tidal regularity: an organisation adopts a governance framework with great ceremony and then never touches it again. The method is installed, the roles are certified, the templates are circulated, the auditors are satisfied — and somewhere in that procession it quietly stops being a way of making decisions and becomes a way of demonstrating that decisions were made. This essay is about that substitution, and about why it is so hard to see while it is happening.
The argument is simple to state and uncomfortable to act upon. The failure of transformation governance is rarely a failure of the framework an organisation chose. It is almost always a failure to adapt it — to fit a general method to a particular enterprise, and then to keep fitting it as that enterprise changes. Adoption is an event with a date and a launch; adaptation is a discipline with no end. We are good at the first and poor at the second, and the same forces that make adoption attractive — the appetite for defensibility, the comfort of a certified method, the sunk cost of a rollout — actively punish the second. What follows examines those forces, takes seriously the strongest case for holding a framework rigidly rather than adapting it, and then draws the distinction that the whole subject turns on: the difference between conforming to a framework and being governed by one. It ends on a proposition that ought to trouble any steering committee that prides itself on stability — that a governance system nobody has ever changed is not mature. It is inert.
The report was green
Picture a programme board of the familiar kind. It meets monthly. The pack runs to forty pages and arrives, as a small act of mercy or self-protection, the evening before. On the summary page the status is green. It has been green for four reporting cycles. The milestones show as met, the risk log is tended, the issues are “being managed”, and the stage-gate that stands between the current phase and the next has been passed, as every previous gate was passed, with a recommendation to proceed.
Everyone in the room knows the programme is in trouble.
They know it in the way that experienced people know things before the instruments admit them. The vendor’s delivery has slipped in substance while holding its dates on paper. The business benefit that justified the whole endeavour — a reduction in the cost of processing, say, of a specified number of pounds per transaction — has quietly detached from the plan and now lives in a later phase that keeps receding. Two of the people who understood the original design have left. And yet the governance runs beautifully. The templates are complete. The gate produced its decision. The audit trail is immaculate. The machine is doing precisely what it was built to do, which turns out to be the problem: it was built to certify motion, not to detect stall.
This is not a story about a bad framework. The framework in the room might be an excellent one — a mainstream programme method, a respectable stage-gate discipline, the kind of structure the Office of Government Commerce and every serious consultancy would recognise and approve. It is a story about what happens to a good framework that is adopted and never adapted: it hardens into a set of motions that are performed faithfully and mean progressively less.
Adoption is an event; adaptation is a discipline
To adopt a framework is to make a decision once. A method is selected, usually a reputable one — PRINCE2 for projects, Managing Successful Programmes for the wider change, a gated model borrowed from product development, a scorecard for the executive view. Training is commissioned. Templates are branded. A programme office is stood up to hold the whole thing together. There is a go-live, and the go-live is real: for a while, the organisation genuinely does more of its governing through the new structure. Adoption has a project plan and a completion date, and it can be declared done.
To adapt a framework is something else entirely, and it has no completion date. Adaptation is the continuous work of asking whether this general method, designed for the average of all organisations and therefore the precise shape of none, actually fits this enterprise, this portfolio, this moment — and of changing it when it does not. It means deciding that a particular gate is theatre for a low-risk workstream and removing it, while sharpening the same gate to genuine severity for the workstream that could sink the business. It means noticing that a reporting cadence designed for a two-year programme is strangling a three-month one. It means rewriting a template because it collects the wrong things, and retiring a committee because the decision it was created to make no longer exists.
Adoption asks, “Are we following the method?” Adaptation asks, “Is the method serving us?” An organisation can answer the first with a confident yes for years while the honest answer to the second slides quietly toward no.
The two questions feel similar and are almost opposites. The first is answerable by inspection: open the templates, check the gates, confirm the roles are filled. It is the question auditors ask, and it is the question a programme office is naturally equipped to answer, because conformance is visible and measurable. The second question cannot be answered by inspection at all. It requires judgement about whether the visible activity is producing better decisions, and judgement of that kind is uncomfortable, contestable, and owned by no one in particular. So we answer the question we can measure and let it stand in for the question that matters — and the substitution is invisible precisely because the measurable question keeps returning a reassuring result.
The forces that reward fidelity and punish adjustment
If adaptation is so obviously valuable, its rarity demands an explanation. It is not laziness, and it is not stupidity — the people running these programmes are neither. The pattern persists because a set of real structural forces makes fidelity to an unaltered framework the rational, defensible, career-safe choice, and makes adaptation look like recklessness. Four of them do most of the work.
- Defensibility. We govern, increasingly, in the shadow of the audit. In the years since the corporate accounting scandals at the start of the decade and the compliance regime that followed them, the unspoken purpose of a great deal of governance has shifted from making the right call to being able to demonstrate that a proper process was followed. An unaltered, brand-name framework is a magnificent instrument of defensibility. If a programme fails, no one is dismissed for having applied PRINCE2 by the book; someone might well be dismissed for having modified it and been wrong. Fidelity transfers responsibility from the individual to the method, and in a blame-conscious culture that transfer is worth more than a better outcome.
- Identity and sunk cost. A framework, once rolled out, becomes something the organisation is. Careers are now attached to it; a cohort has been certified in it; a programme office exists to administer it and would have less to administer if it were pared back. The cost of the rollout — the training, the tooling, the year of change — is sunk, but sunk cost casts a long shadow over the willingness to change what was bought. To question the framework starts to feel like questioning the people who championed it, and organisations are exquisitely sensitive to that.
- The comfort of conformance. Following a defined method is genuinely easier than exercising judgement, and not only because it is less work. It is less exposed. When the template tells you what to collect and the gate tells you what to ask, the anxiety of deciding what ought to be collected and asked is lifted. Conformance offers the profound relief of a clear conscience: I did what the method required. That relief is real, and it is addictive, and it quietly removes the discomfort that would otherwise prompt someone to notice the method had stopped working.
- No owner of the question. Every part of a governance system has an owner except the system itself. The gate has a chair, the template has an author, the programme office has a head — but the question of whether the whole apparatus is still fit for purpose belongs to no role, appears on no plan, and is measured by no report. What no one owns, no one does. The framework is therefore never adapted, not because anyone decided against it, but because deciding for it was never anybody’s job.
Set these four forces beside one another and the outcome is over-determined. Fidelity is safe, visible, rewarded, and owned; adaptation is exposed, invisible, unrewarded, and owned by no one. That a good framework ossifies is not a surprise. It is the predictable result of the incentives we have built around it.
The strongest case for holding the line
It would be too easy to leave it there, as though the only reason anyone resists adaptation were timidity or bureaucratic self-interest. There is a serious argument on the other side, and any honest treatment has to meet it at full strength rather than in caricature.
The argument runs like this. In a large organisation, consistency is not a bureaucratic fetish; it is a genuine good, and often an underrated one. When every programme governs itself differently, the executive loses the one thing a portfolio view is supposed to give them — the ability to compare. A green here no longer means what a green there means. Assurance cannot function, because there is no common standard to assure against. Worse, “adaptation” is very often the flag under which discipline is quietly abandoned: the gate that gets “tailored out” is usually the inconvenient one, the review that gets “streamlined” is the one that would have asked the hard question. Give people permission to adapt the framework and a certain number of them will use it to escape the parts of the framework that exist precisely to constrain them. A method half-followed can be worse than either extreme, because it offers the reassurance of governance without the substance. On this view, rigidity is not a failure mode at all. It is a defence — against local self-interest, against the erosion of standards, against the human tendency to remove exactly the check that was about to catch us.
This argument is correct in its facts and wrong in its conclusion, and seeing why is the crux of the whole matter.
It is correct that undisciplined, self-serving “adaptation” is a real and common failure — arguably more common than the rigidity this essay is warning against. It is correct that consistency has genuine value and that assurance needs a common standard. Where it goes wrong is in assuming that the alternative to rigidity is licence: that once you permit the framework to be changed, you have no principled way to stop it being gutted. That assumption is what makes rigidity look safe. But it is false, and it is false because it confuses two different things that a framework contains — its mechanisms and its intent.
Governance as a decision engine, not a reporting rite
Every governance framework is really two things wearing one name. There is the intent: the underlying questions the framework exists to force an organisation to confront. Is this still worth doing? Can we actually deliver it? Are the benefits real and are we on track to realise them? What would make us stop? And there is the mechanism: the specific gates, templates, cadences, committees, and status codes through which those questions are, in theory, asked.
The intent is close to universal and rarely needs changing. The mechanism is local, perishable, and needs changing constantly — because a mechanism that asks the right question in one context asks nothing in another. A stage-gate is a mechanism for forcing the question should we continue? But a gate that has been held forty times and has never once produced any answer other than “proceed” is no longer asking that question. It has become a mechanism for producing the answer “proceed” — a ritual that confers permission rather than a test that might withhold it. Its intent has drained out of it entirely, and its faithful, unaltered repetition is the very thing hiding the fact.
“A gate that cannot say no is not a gate. It is a turnstile with a ceremony attached.”
This is where the distinction dissolves the objection. Disciplined adaptation does not touch the intent; it defends the intent by repairing the mechanism when the mechanism stops serving it. The undisciplined “adaptation” that the objection rightly fears is the opposite: it removes a mechanism because the intent is inconvenient — it tailors out the gate precisely because the gate might say no. The two are not points on a spectrum from which any movement is dangerous. They are opposite in direction. One changes the mechanism to preserve the question; the other suppresses the question by removing the mechanism. A governance system that knows the difference can adapt freely and lose nothing, because it holds the intent fixed and treats every mechanism as provisional.
Consider what that looks like in the concrete. A programme I was once close to ran a monthly board with a single overall status, reported as red, amber, or green. For most of its life the status was green while the one figure that actually governed its worth — the realised benefit against the forecast benefit — sat on a later page, unlinked to the colour on the front. The mechanism (a RAG status) was intact and faithfully produced; the intent (are the benefits real and on track?) had quietly left the building. The adaptation, when it finally came, changed almost nothing structurally and everything substantively: the front-page status was redefined so that it could not be green while realised benefit trailed forecast by more than a stated margin, and the gate was given a standing instruction that a phase could not pass while that gap was open without a named executive signing personally for the exception. No new committee. No new template of consequence. The same framework — but the mechanism was re-coupled to the intent it had drifted away from. Within two cycles the status told the truth, and the conversations in the room changed from narrating the plan to defending the benefit. The framework had not been diluted. It had been switched back on.
| Framework as adopted | Framework as adapted |
|---|---|
| The gate asks the questions the template lists | The gate asks the questions this decision actually turns on |
| Status reports conformance to plan | Status reports distance from the outcome |
| Consistency means identical mechanisms everywhere | Consistency means the same intent enforced everywhere |
| Success is a passed review | Success is a better decision, sometimes a decision to stop |
| Changing the framework is a risk to be avoided | Not changing the framework is the risk |
Note what the right-hand column preserves. It is not less rigorous than the left; in most respects it is more so. It keeps the intent absolutely fixed — indeed it enforces the intent more severely than the unaltered framework did, because it has stopped letting a passed gate stand in for a sound decision. What it holds lightly is only the mechanism, and it holds the mechanism lightly precisely so that it can hold the intent tightly. That is the answer to the objection. Consistency is preserved at the level that matters — every programme is still made to confront the same fundamental questions — while the specific instruments are allowed to be whatever actually forces those questions in each context. Rigidity offers consistency of form and surrenders consistency of substance, which is the more valuable of the two and the harder to see slipping.
The living framework
There is a tempting conclusion available here, and it is the wrong one. The wrong conclusion is that organisations should adopt frameworks more loosely, hold them more sceptically, believe in them less. That is not it at all. The organisations that govern well are not the ones that believe in their frameworks least; they are the ones that take the intent of governance with deadly seriousness and hold the mechanism with an open hand. They are, if anything, more demanding of their governance than the faithful conformer is, because they refuse to accept a passed gate as proof of a sound decision. Belief in the intent and lightness with the mechanism are not in tension. They are the same disposition.
What that disposition requires, practically, is almost embarrassingly modest, and its modesty is why it is so rarely done. It requires that someone be made to own the question no one owns — that the fitness of the governance system itself be put on a plan, given a name, and reviewed with the same seriousness the system brings to everything else. It requires that at least once in the life of a programme, and ideally on a rhythm, the board turns its instruments on itself and asks not “did we follow the framework?” but “did the framework help us decide anything we would not otherwise have decided, and where did it merely certify what we were going to do regardless?” A gate that has never changed a decision is not a stable gate. It is a dead one, and the discipline is simply the willingness to notice.
The maturity of a governance system is not measured by fidelity to the framework it adopted. It is measured by the quality of the decisions the framework produces — and the surest sign that it is still alive is that the organisation has been willing, more than once, to change it.
So the test to carry away is not whether your governance is being followed. It almost certainly is; conformance is the easy part and the part that flatters us. The test is whether it is still deciding anything — whether, somewhere in all that faithful motion, a gate could still say no, a status could still tell an unwelcome truth, a review could still stop a programme that everyone had already agreed to continue. If you cannot remember the last time your governance changed a decision, it has become the machinery of certification, and no amount of fidelity will bring it back to life. Only adaptation will. A framework you have never adapted is not one you have mastered. It is one that has quietly stopped working while you were busy following it.