Somebody in your organisation has already fixed it. They just have not told you.
Somewhere in most organisations right now, someone has quietly rebuilt a piece of work that was never officially theirs to fix, cut a four-hour task to forty minutes, and mentioned it to no one.
Not because they are hiding it. Because they are not sure whether they are allowed.
That person is doing exactly what the strategy says the business needs. The organisation just does not know, so it cannot learn from it, scale it, or protect the person doing it in the next restructure.
This article is about why that happens, and what four decades of research on motivation, psychological safety and organisational design say about changing it. It is written for senior leaders in mid and large companies who suspect the capability they keep paying consultants to build is already in the building.
If the person who fixed something last week told you tomorrow, what would actually happen next?
Somewhere in your business right now, a person you would struggle to name has quietly rebuilt a piece of work that was never officially theirs to fix, cut a four-hour task to forty minutes, and mentioned it to no one. Not because they are hiding it. Because they are not sure whether they are allowed.
That person is doing exactly what every strategy document says the organisation needs. They noticed something was not working, took responsibility for improving it, and did it inside the boundaries of their actual role. The only thing missing is that the organisation does not know, and so cannot learn from it, scale it, or protect the person doing it from the next round of restructuring.
This is the territory of intrapreneurship: entrepreneurial behaviour exercised from inside an existing organisation rather than by leaving to start a new one. It is not a new idea. What is new is how much of it is now happening in private, and how expensive that privacy has become.
Whatever sector you are in, the shape of the pressure is familiar. Customers expect more, faster and for less. Costs have risen faster than pricing power. Technology, and generative AI in particular, is moving faster than policy, procurement or process can absorb it. And several rounds of cost programmes have removed layers and roles, leaving fewer people to carry the same expectations.
In this environment, the organisations that cope well are rarely the ones with the most ambitious strategies. They are the ones with enough people inside the system who notice what is not working and act on it without waiting for a mandate. The question for a senior leader is not whether you have those people. You almost certainly do. The question is whether your organisation lets them be useful.
Gifford Pinchot coined the term in the mid-1980s, in Intrapreneuring (1985), to describe employees who take hands-on responsibility for creating innovation inside a company rather than leaving to found their own. Peter Drucker was writing about the same shift in large organisations at the same time, in Innovation and Entrepreneurship (1985). Pinchot's definition is worth holding onto: taking advantage of a new opportunity and creating economic value within the company itself. The research since has tied the behaviour to growth, competitiveness and firm survival, and more precisely to organisational effectiveness and value creation rather than simply to output.
That distinction, between activity and value, is the whole matter. Most organisations are full of busy people. Fewer are full of people asking whether the busyness produces anything that matters. Think about your last budget round. The default response to a tighter number is to do the same things with fewer people. The intrapreneurial response is to ask which of the things should stop. Those are not the same conversation, and most organisations only know how to have the first one.
Leaders can mandate a process change. They cannot mandate ownership, and the attempt tends to produce the opposite. This is where organisational psychology stops being decorative.
Edward Deci and Richard Ryan's self-determination theory, developed from the 1980s onward, offers one of the most robust and thoroughly tested explanations of what drives this kind of behaviour. People are most motivated, creative and persistent when three needs are met: autonomy, a sense of choice over how work is done; competence, a sense of being capable and improving; and relatedness, a sense of connection to others and to purpose. Where these are suppressed, typically by tight control, heavy oversight or a culture built on checking rather than trust, initiative recedes and people default to the minimum required.
Cost pressure does this by stealth. When budgets tighten and headcount falls, the organisational reflex is to centralise, add sign-offs and manage risk by control. The people who respond worst to that reflex are exactly the ones you most need to keep: the ones who look at a process and say "this step exists because of a rule that no longer applies". They are also, in a loose labour market for good people, the ones most able to leave. The capability walks out the door at the moment it is most valuable, and the exit interview records "wanted a new challenge".
Hybrid work has quietly eroded the third leg of the theory. People now have more autonomy over where they work and weaker connection to purpose and to each other. Relatedness is what makes someone willing to put their neck out for an idea. Without it, autonomy just means working alone.
Teresa Amabile's componential theory of creativity (1983, refined 1996) found that people generate more original and useful ideas when they have meaningful freedom in how they approach a problem, when they are intrinsically motivated by the work, and when the climate actively supports risk-taking rather than merely tolerating it. Her later work with Steven Kramer on the progress principle (The Progress Principle, 2011) found that the single strongest driver of motivation on any given day is the sense of making meaningful progress on work that matters. It outranks recognition and it outranks incentives.
The implication for an intrapreneurship effort is direct: small, visible wins matter more than large, distant ones. A two-week pilot that removes one unnecessary approval step teaches more, and builds more confidence, than a business case that takes six months to approve. The hybrid complication is the word visible. A small win in a distributed team is one nobody was in the room to notice. If your organisation has no mechanism for making small progress seen, the progress principle is working against you rather than for you.
Amy Edmondson's research on psychological safety, from her 1999 study of hospital teams through The Fearless Organization (2018), explains why so many good ideas inside organisations never surface. Teams only raise concerns, admit mistakes and propose unconventional ideas when they believe they can do so without being embarrassed, marginalised or punished. Without that belief, intrapreneurial behaviour is privately present and publicly absent. People notice the problem. They just do not say so.
Return to the person from the opening. There is a fair chance the tool they used was one nobody issued them. Shadow use of generative AI is now widespread in most organisations, and it is intrapreneurship without psychological safety. The experimentation is happening. The learning is not, because the experimenter cannot tell whether the response to "I automated this" will be curiosity or a reprimand. This is not an argument against rules. Intrapreneurs need the boundaries to be clear, because clear boundaries are what make an experiment safe to run and safe to report. Every organisation that has issued a restrictive AI policy without a matching invitation to experiment inside it has, in effect, instructed its most inventive people to keep their inventions to themselves.
There is a generational version of the same dynamic. Younger cohorts are, on the whole, less patient with "that is how we do it here" and more likely to leave than to lobby. That is intrapreneurial energy the organisation is losing rather than harnessing, and it is often lost in the first year, before anyone senior has met the person.
Rosabeth Moss Kanter's study of innovation inside large companies, The Change Masters (1983), found that innovation clusters in organisations with integrative structures: fewer rigid silos, more cross-functional contact, and boundaries between departments treated as porous rather than fixed. Segmented, hierarchical organisations produced far less innovation regardless of how talented the people in them were. This reframes intrapreneurship as a design question, not only a mindset question.
It also explains why delayering cuts both ways. Many organisations have removed management layers in the last few years. Fewer layers can mean shorter approval chains and more porous boundaries, which is what Kanter found supports innovation. But it can also mean the remaining managers are stretched so thin that they default to control because control is faster than coaching. Delayering without redesigning decision rights is a structural change that removes the people who used to say yes without adding anyone who can.
David Teece's work on dynamic capabilities (1997, developed in 2007) adds the strategic lens. An organisation's long-term advantage rests less on what it holds at any one point and more on its capacity to sense change, seize opportunities by reconfiguring how it works, and continuously transform. Intrapreneurs are the human mechanism through which sensing and seizing happen at ground level. The value from AI in most businesses is not coming from the AI strategy. It is coming from people close to the work who rebuild a process, drop a reporting step and then ask what to do with the hours. A strategy can name the opportunity. Only people with the standing and confidence to act can take it.
Donella Meadows argued, in Leverage Points (1999), that most attempts to fix organisational problems intervene at the weakest point: adjusting a number, a budget or a target, when the real leverage sits upstream in the rules, incentives, information flows and mental models producing the behaviour. This is why so many improvement efforts deliver a short uplift that fades. They treat a symptom, not the structure generating it.
Ronald Heifetz's distinction between technical and adaptive challenges (Leadership Without Easy Answers, 1994) sharpens the point. Technical problems have known solutions and can be delegated to an expert or a process. Adaptive challenges require the people experiencing the problem to change their own assumptions and ways of working, and cannot be solved by authority alone. Almost everything on a senior leader's agenda now is adaptive: AI adoption, cost pressure alongside rising expectations, a workforce with different terms of engagement. This is precisely the terrain where intrapreneurs operate best, because they are close enough to the problem to see the real pattern and motivated enough to test a different response.
Bringing the strands together:
- Start with mindset, not tools. Before introducing frameworks or platforms, help people see the difference between reactive delivery and proactive value creation, and where their own habits keep them in the former. The AI tools are already in the building. The mindset is the constraint.
- Protect small experiments. Following Amabile, small, safe, fast tests build more sustained motivation than one large sanctioned programme. Make progress visible on purpose, especially in distributed teams, because nobody will see it by accident.
- Build psychological safety deliberately. Leaders have to visibly reward the surfacing of problems and the reporting of failed experiments, not only successes, or Edmondson's dynamic reasserts itself. The fastest test is what happens the first time someone says "I automated this without asking".
- Redesign structure alongside mindset. Kanter's research is a caution against running a capability programme inside a structure that will not let the capability be used. If you have delayered, decision rights have to be redrawn, not assumed.
- Diagnose the pattern before designing the fix. Applying Meadows, ask what structure is producing a recurring problem before proposing a solution. Ask which stories ("we tried that once", "the board will never wear it") are doing the work of a rule.
- Build the ability to read the wider ecosystem. Intrapreneurs need to understand who the customer actually is, what stakeholders across and beyond the organisation care about, and where an idea will find support rather than resistance. Kanter and Teece both point the same way: people who can see across boundaries are the ones who spot where value is being created or lost.
- Give people the skills to navigate internal politics. Good ideas fail for social reasons as often as technical ones. Stakeholder mapping, coalition building, timing and framing are core intrapreneurial skills, because they are what let a well-designed experiment survive contact with the rest of the organisation.
The wider case
None of this is about entrepreneurship in the start-up sense. It is about resilience. Companies that build real intrapreneurial capacity reduce their dependence on any single leader or star performer, because the capability to notice, test and improve is embedded in ordinary roles rather than concentrated at the top. That is a more durable form of change than any transformation programme, and it is available to organisations of almost any size or budget, because the conditions that produce it, autonomy, mastery, psychological safety, workable structure and honest diagnosis of the system, sit within leadership's control regardless of how constrained the wider environment is.
The organisations that manage the coming decade well are unlikely to be the ones with the most ambitious strategies. They will be the ones where enough people, at enough levels, feel able to treat the organisation as something they are responsible for improving rather than merely something they work inside. So a question worth asking yourself: if the person who quietly fixed something last week told you about it tomorrow, what would actually happen next?