Making Innovation Stick (Revisited)
Last year, I wrote a piece on Making innovation stick: lessons from the frontlines of public sector innovation.
It was grounded in practice. Born from hard-won lessons inside systems trying, and often struggling, to change.
Coming back to it now, I think it was pointing in the right direction. But it didn’t quite get to the heart of the matter.
That’s a familiar pattern in innovation work: a pilot shows promise, the early signals are good, people lean in, a report is written, a case study polished. There’s talk of scaling. And then… nothing much happens.
Or, perhaps worse, something does happen. The idea is rolled out quickly, broadly, confidently. For a while it looks like progress. And then, quietly, it fades. Workarounds creep back in. Old habits reassert themselves. The system absorbs the change and carries on much as before.
We tend to treat this as a failure of execution, a gap between ambition and delivery.
But more often, it’s something else. Not an innovation problem, a persistence problem.
The myth of transfer
Much of how we think about innovation rests on a simple assumption: that good ideas can be moved. If something works in one place, we should be able to replicate it elsewhere. Package it, transfer it, scale it.
But systems don’t work like that.
What looks like a successful intervention is always entangled with the context it emerged from: the relationships that held it, the informal practices that sustained it, the local workarounds that made it viable. Strip those away, and what remains is often just the visible surface.
We try to move the artefact, but what actually mattered was the pattern of practice underneath it. And patterns don’t transfer cleanly, they have to be grown into place.
What “sticking” really means
If we take this seriously, then success starts to look different.
Something has stuck when:
It survives beyond the enthusiasm of its original champions
It becomes part of how work is actually done, not just how it is described
It continues under pressure, even when attention shifts elsewhere
You can think of this as three stages:
Adoption: people try it
Embedding: it fits into everyday workflows
Endurance: it holds under real-world conditions
Most innovation efforts make it through the first stage, some reach the second, but very few achieve the third. And it’s in that gap that most of the value is lost.
Why things don’t stick
In the last piece, I argued that our fixation on speed is part of the problem. The pressure to move quickly leaves little room for the slower work of adaptation.
But there are other dynamics at play:
We over-design solutions, leaving little space for local ownership.
We prioritise visibility over viability.
We treat resistance as something to overcome, rather than something to learn from.
And often, we underestimate the system itself. Systems are not passive recipients of innovation. They respond. They adapt. Sometimes they resist outright. Sometimes they absorb change just enough to neutralise it. Either way, the result is the same: the appearance of progress, without the substance of it.
The craft of making it stick
If innovation that lasts looks less like a breakthrough and more like a slow embedding, then the question becomes: what kind of work makes that possible?
This is where innovation starts to look less like strategy and more like craft. Not craft in the nostalgic sense, craft as in the disciplined, situated practice of shaping something with the material you’re working in. In this case, the material is the system itself: its routines, relationships, incentives, histories.
And like any craft, much of this work relies on forms of knowledge that are hard to codify. Michael Polanyi described this as tacit knowledge: the idea that we know more than we can tell. A skilled practitioner can sense where resistance will emerge, recognise when a small adjustment will unlock movement, feel when something is slightly off before it becomes visible in the data. Human systems are no different.
You can map stakeholders, draw processes, model incentives. All are useful, but none of it quite captures how a system actually behaves when you try to change it. That only becomes visible through immersion, through working inside it, through paying attention to the small signals most frameworks filter out.
This is why so many well-designed innovations fail to take hold. They are built on explicit knowledge alone: what can be written down, transferred, scaled. But they lack sensitivity to the tacit layer where real adoption happens.
Making innovation stick requires working at that level. It means:
Starting with fit, not novelty: understanding what this system is ready to absorb, not what looks impressive from the outside
Building with insiders, not for them: because tacit knowledge sits with the people who live the system every day
Designing for friction: treat resistance as information, not failure. Friction reveals where the system’s grain runs
Staying long enough to learn the grain: not just diagnosing from a distance, but shaping and reshaping in situ
Over time, this kind of work creates something different from a successful pilot, it creates alignment between the intervention and the system itself. The change no longer feels imported, it begins to feel natural. And that is when innovation starts to stick.
A different way of thinking about innovation
All of this points to a deeper shift. Innovation isn’t about introducing change, it’s about skillfully shaping how systems evolve so they can do what they are meant to do.
That’s slower work, more relational and less visible in the short term. But it is also more durable, because instead of trying to impose change from the outside, it works with the system itself. Its constraints, its capabilities, its direction of travel. And over time, that changes not just what the system does, but how it functions.
What comes next?
Even this kind of careful, craft-based work doesn’t guarantee success, because some systems are not just resistant to change, they are actively structured to reject it.
In the next piece, I’ll explore that dynamic in more detail: how systems develop their own immune responses to innovation, and what that means for anyone trying to work within them.
