Why Schools Don't Have an Implementation Problem

There is a familiar moment in school improvement. An initiative has been introduced. Professional learning has been provided. Resources have been purchased. Leaders have communicated expectations. Time has passed.

And it is not working.

Perhaps implementation is inconsistent. Some teachers have embraced the new approach while others use it occasionally. Walk through classrooms and the variation is obvious. Ask leaders what has happened and eventually somebody says it.

“We have an implementation problem.”

They may be completely right.

Schools have implementation problems. Good ideas can be poorly introduced. People can receive insufficient training. Leadership attention can drift. Resources can disappear. Expectations can remain vague. An initiative that initially generated enthusiasm can slowly be displaced by everything else a school is required to do.

So the title of this essay is deliberately unfair. Of course schools have implementation problems.

The more interesting question is whether implementation problem sometimes becomes the name we give to a failure because implementation is where the failure finally becomes visible.

I have seen what that looks like.

Imagine a school that has undergone significant pedagogical transformation. It has moved away from a highly traditional model of teaching and invested deeply in inquiry, conceptual learning, student agency and sophisticated curriculum design. New pedagogical leaders have been appointed. Teachers have undertaken substantial professional learning. Practice has evolved. The organisation has developed considerably greater capacity to think about teaching and learning.

Then a concern emerges: student engagement needs improvement.

That might be an entirely legitimate concern. Engagement matters enormously, and even highly accomplished schools should remain capable of questioning their practice. But before a solution is chosen, there is a more fundamental question.

What evidence tells us that this is the problem?

Not whether engagement matters. Not whether an intervention exists that claims to improve it. Not whether another respected school is using it. What, specifically, is happening here?

Where is it happening, and for whom? Under what conditions? Is the problem evident across the whole school, or in particular parts of it? Does engagement mean the same thing in a play based early childhood environment as it does in a secondary mathematics classroom? What is already working? What do the people closest to teaching and learning understand about the issue?

These are not implementation questions. They precede implementation.

In the situation I experienced, a whole school approach was introduced. External expertise was brought in. Considerable money and professional learning time were invested. Teachers were trained in particular engagement strategies and expected to demonstrate them in practice. Leaders incorporated the same protocols into meetings. Lessons were observed. From the outside, many of the recognisable components of implementation were present.

But something important was missing.

The people responsible for pedagogical leadership had not been meaningfully involved in identifying the problem or selecting the response. There was little shared conviction that the intervention addressed a demonstrated need. Teachers were already working within a sophisticated pedagogical model, and some experienced the new strategies not as an extension of that practice but as an additional layer sitting awkwardly on top of it. Nothing significant had been removed to make room.

Eventually the predictable question emerged: why isn't this sticking?

The easiest answer was implementation. Perhaps people needed another session. Perhaps leaders needed to model the strategies more consistently. Perhaps classroom observations needed greater focus, or expectations needed strengthening. Any of those things might have improved fidelity.

None answered the question of whether the school should have been implementing this particular solution, in this particular way, in the first place.

That distinction is not merely philosophical. Implementation research itself gives us a much more sophisticated vocabulary than simply asking whether people are “doing it”. Proctor and colleagues distinguish between outcomes such as acceptability, adoption, appropriateness, feasibility, fidelity and sustainability. The distinction between appropriateness and fidelity is particularly useful. Fidelity asks whether something is being implemented as intended. Appropriateness asks whether it is actually perceived as relevant and compatible with the setting, people and problem.

Those are very different questions.

A school can therefore become increasingly effective at ensuring that people perform an agreed practice without ever resolving whether that practice is the right response to the problem it faces. If the people expected to enact a change do not perceive it as relevant to the problem they encounter, that is not simply an attitude to overcome once implementation begins. If an intervention sits awkwardly alongside existing practice, additional training may improve people's capacity to perform it without improving its fit. And if the problem itself has been poorly specified, greater fidelity may simply result in people implementing the wrong response more consistently.

This does not mean practitioner scepticism proves leadership wrong. That would be an equally dangerous simplification.

People resist worthwhile change. Familiar practice can feel safer than better practice. Strong evidence can challenge deeply established professional beliefs. Some innovations require persistence before their benefits become visible, and leadership sometimes has to hold direction through discomfort rather than interpreting every objection as evidence that an initiative lacks value.

The question is not whether people like the change. It is whether the organisation has done enough work to understand what their response to it means.

This is one reason contemporary implementation research takes context so seriously. Frameworks such as the Consolidated Framework for Implementation Research direct attention not only towards the people expected to enact an innovation, but towards the innovation itself, the setting in which it is being introduced and the process through which implementation occurs. Educational implementation guidance has increasingly moved in the same direction, recognising that implementation depends upon behaviours, context and careful selection, not simply the quality of the programme being introduced.

That last idea, selection, deserves more attention than it often receives.

Schools can become extraordinarily sophisticated at implementing solutions and surprisingly imprecise about identifying problems. A new programme arrives with materials, professional learning, timelines and expertise. It gives an organisation something concrete to do. Problem diagnosis is less satisfying. It requires uncertainty. It may reveal that different parts of a school are experiencing different problems, that something already working well should be strengthened rather than replaced, or that the issue we thought belonged to teaching actually belongs somewhere else entirely.

Solutions create a sense of movement. Diagnosis can feel slower. Yet implementation eventually inherits whatever the diagnosis leaves unresolved.

If a school has not established why a change is necessary, implementation inherits ambiguity of purpose. If people hold different understandings of the problem, it inherits fragmented meaning. If an initiative is added to an already overloaded environment without anything being removed, it inherits competing attention. If leaders privately question a decision but publicly present certainty, it inherits fragile belief. If practitioners with deep knowledge of the work were excluded from shaping the response, implementation inherits the consequences of knowledge that never entered the decision.

Then, six months later, we measure what teachers are doing.

That is why implementation can become such a seductive place to locate responsibility. It is visible. A leader can observe whether a strategy appears in a classroom. A checklist can record whether agreed practices are present. Participation in training can be counted. Fidelity can be measured.

The conditions that preceded implementation are harder to see. We cannot so easily measure how much belief existed before launch, how well the solution fitted existing practice, whose knowledge influenced the decision or which assumptions went untested. It is harder to see the competing commitments already consuming people's attention, whether everyone understood the purpose in the same way, or whether the initiative was solving a problem practitioners actually recognised.

Implementation brings those conditions into contact with reality. And reality can be remarkably diagnostic.

Implementation failure is evidence. The mistake is assuming it is evidence only about implementation.

This changes how we interpret inconsistency.

Imagine an observer enters a classroom and sees the new protocol performed exactly as expected. The teacher uses the required strategy. The observer leaves. The following week, the strategy disappears.

One interpretation is obvious: the teacher is not implementing consistently.

True.

But the inconsistency itself does not tell us why. The teacher may lack confidence and need coaching. Workload may be making sustained implementation unrealistic. The strategy may not yet have become habitual. Expectations may be unclear or leadership attention inconsistent. Equally, the teacher may be perfectly capable of performing the strategy and have made a professional judgement that it adds little value to the learning taking place.

The visible behaviour is the same. The conditions producing it may be entirely different. More training will help some of those conditions. Greater accountability may address others. Neither is a universal answer.

This is why compliance can be so misleading. Visible compliance tells us that someone can perform an expected behaviour under particular conditions. It does not necessarily tell us that they understand its purpose, believe in its value, can integrate it coherently with existing practice or intend to continue when scrutiny disappears.

In the school I described, teachers could demonstrate the required engagement strategies. That was never the deepest problem. The difficulty was that many of the people responsible for carrying the work did not believe the addition represented an improvement on the sophisticated practice already developing around them.

Implementation was therefore being asked to produce a conviction that the decision making process itself had never established.

Of course, conviction does not always have to exist before implementation begins. Understanding can deepen through experience. People can change their minds when they see an approach working. Strong implementation can build confidence in something that initially felt unfamiliar or uncomfortable.

But there is a difference between belief developing through meaningful engagement with change and belief being assumed because the decision to change has already been made.

This becomes particularly interesting in organisations that value agency.

Schools may invest deeply in student agency, teacher agency, inquiry and professional judgement while implementing organisational change through processes that offer relatively little of those things to the adults involved. The contradiction may never be intentional. Leadership may simply be trying to create coherence. A whole school initiative may genuinely require common expectations, and consultation cannot mean every individual receives veto power. At some point, decisions have to be made.

But if the philosophy of learning says that meaning is constructed, context matters, inquiry deepens understanding and agency strengthens ownership, there is at least a question worth asking when organisational change is introduced through a process that largely says: here is the solution, learn it, use it and show us that you are using it.

The method of implementation can quietly contradict the beliefs the organisation is trying to implement.

There is another complication, and it is one organisations may find even harder to entertain.

Sometimes the idea itself is not good enough.

Implementation discourse can become so focused on how to embed an intervention that it inadvertently grants the intervention immunity from further scrutiny. Once significant money, time and leadership credibility have been invested, questioning the initiative can begin to feel like undermining implementation. Yet a decision does not become better because it has become expensive to reverse.

Implementation should not require organisations to stop evaluating the thing being implemented.

The distinction between fidelity and appropriateness becomes useful again here. A school could theoretically achieve high fidelity to an intervention that was poorly matched to its needs. Everyone could do exactly what was expected of them, exactly as intended, and the organisation might still have implemented the wrong response to the wrong problem.

That would be successful implementation of a questionable decision.

Which takes us back to the hypothetical principal.

“The idea is sound. The evidence is strong. We communicated it clearly. We trained everyone. Six months later, half the school still isn't doing it properly. How can you tell me we don't have an implementation problem?”

I wouldn't.

I would say: You might. But before we fix it, I want to know what your implementation problem is telling us.

I would want to understand the evidence that established the original problem and why this particular intervention was selected over the alternatives. I would want to know where it fits the existing work and where it competes with it, what people understood before implementation began, what was removed to create capacity, and what the practitioners closest to the work said before the decision and are saying now.

Then I would want to see the implementation data.

Perhaps the answer really is better implementation. Perhaps the idea is appropriate, feasible and worthwhile, but people need greater support. Leaders may need to protect time, strengthen coaching, improve resources or maintain attention long enough for new practice to become established.

If so, do that.

But perhaps implementation is revealing something upstream that another training day will never repair.

That possibility changes the leadership question. Instead of moving immediately to How do we make people do this?, we can ask:

What is this difficulty revealing?

That is a more demanding question because the answer may not sit with the people whose practice is easiest to observe. It may sit in the original diagnosis, the assumptions that shaped the solution, the organisational conditions into which it was introduced, the amount already on people's plates, the way the decision was made, or the intervention itself.

Plans can look beautifully coherent on paper. Presentations can make purpose sound obvious. Leadership teams can reach agreement and professional learning can generate genuine enthusiasm. Eventually, though, the idea has to meet the ordinary complexity of Monday morning. It has to coexist with the teacher's learners, professional judgement, existing curriculum, timetable constraints, accumulated workload, established habits and everything else the organisation has previously asked people to care about.

Implementation is where aspiration encounters conditions.

Sometimes those conditions need to change. Sometimes implementation itself needs to improve. Sometimes people need greater support or clearer expectations. And sometimes the organisation needs the courage to reconsider the decision that sent everyone there.

Schools do have implementation problems. But the phrase becomes dangerous when it allows us to begin our diagnosis at the place where the symptoms appeared.

The next time an initiative falters, by all means ask whether it has been implemented well. Then ask another question.

What did we ask implementation to carry that should have been resolved before implementation began?

Because sometimes implementation is not the place where improvement fails. It is simply the place where the organisation can no longer avoid seeing why.