There is a type of business that is always in motion.

New campaigns every quarter. New tools adopted and then quietly abandoned. New hires brought in to fix problems that the last hire was supposed to fix. New strategies announced with energy and conviction, followed — twelve months later — by a new strategy announced with equal energy and conviction.

From the outside, this business looks active. It is responsive to change, willing to experiment, not stuck in its ways. These are qualities that management literature has celebrated for decades. And yet the business does not move forward in any meaningful sense. Its position in the market is roughly what it was three years ago. The same conversations happen in leadership meetings. The same problems recur.

This is not a failure of ambition or effort. It is a failure of accumulation.

The Difference Between Experimenting and Learning

The culture of experimentation has become one of the dominant orthodoxies of modern business strategy. Eric Ries's The Lean Startup, published in 2011, introduced the concept of the minimum viable product and the build-measure-learn loop to a mainstream audience and produced a generation of founders and executives who are comfortable with the idea of running experiments.¹ This is, on balance, a good thing.

But experimentation without a learning architecture is not a growth strategy. It is organised trial and error — and the emphasis, in many businesses, has fallen far more heavily on the trial than on the error.

The critical requirement of the build-measure-learn framework — the part that is most often skipped — is the learn step. Not just absorbing that something didn't work, but understanding why it didn't work, what that reveals about the underlying assumption, and how the next experiment should be designed differently as a result. Without that step, the loop doesn't close. Each experiment is essentially independent. The business generates activity but not knowledge.

The psychologist and organisational theorist Chris Argyris spent decades studying how organisations learn — or fail to. His distinction between single-loop learning (adjusting actions when outcomes don't match expectations) and double-loop learning (questioning the assumptions and mental models that produced the actions in the first place) is one of the most practically useful frameworks in management theory.² Single-loop learners try different things. Double-loop learners understand why things work or don't and redesign from that understanding. Most businesses that are perpetually in motion are single-loop learners: they change tactics without interrogating strategy.

Reset Culture vs. Build Culture

There is a deeper pattern underneath the experimentation problem, and it is cultural.

Some organisations are wired to build — to treat each initiative as a deposit into a cumulative asset. The campaign teaches them something about their customer. The new sales approach reveals something about how the market evaluates them. The product iteration eliminates a friction point that was limiting retention. Each cycle leaves the business in a slightly better position than the previous one, even when the immediate result was disappointing.

Other organisations are wired to reset — to treat failure (or even modest success) as a signal to start over. The campaign didn't work, so the strategy is discarded. The new hire didn't deliver quickly enough, so the function is restructured. The new tool was supposed to solve the problem, but it didn't, so a different tool is evaluated. There is no accumulation because nothing is allowed to compound.

The reset impulse is not always wrong. Sometimes the strategy genuinely needs to change. But in organisations where resetting is the default response to difficulty, it often serves a different function: it avoids the harder work of examining whether the problem is the approach or the execution, and whether what looks like failure is actually a learning that, if properly absorbed, would produce a different result the next time.

The Harvard Business School professor Amy Edmondson, in her extensive research on how teams learn from failure, found that organisations with a reset culture systematically underinvest in what she calls productive failure analysis — the disciplined examination of what went wrong and why.³ In these organisations, failed initiatives are written off rather than mined for insight. The cost is invisible in any single instance. Accumulated over years, it represents an enormous loss of compounding knowledge.

The Seduction of New

There is a cognitive dimension to this pattern that is worth being honest about.

Novelty is intrinsically motivating. Starting something new feels like progress, even when it isn't. The blank page, the new initiative, the fresh start — these trigger a cognitive response that sustained, grinding execution on a partially-working idea does not. Neuroscience research on reward systems has consistently shown that the anticipation of a new outcome activates dopamine pathways more strongly than the continuation of a familiar one.⁴

In a business context, this means that the decision to try something new is often emotionally easier than the decision to stay with something difficult. A new marketing strategy is more exciting than a careful post-mortem on why the last one underdelivered. A new CRM is more energising than a disciplined effort to improve adoption of the current one.

Leadership teams are not immune to this. In fact, the higher the pressure to demonstrate responsiveness and decisive action, the stronger the pull toward visible new initiatives — because these signal action in a way that the patient, unglamorous work of fixing and improving existing systems does not.

The strategist Roger Martin has written about this as the action bias — the tendency of organisations to prefer doing something new over doing something better.⁵ It is reinforced by every quarterly review where leadership is expected to announce what has changed and what is next, rather than what has been learned and how it changes what gets built.

What Accumulation Requires

Businesses that genuinely move forward — not just in activity but in position — share a structural characteristic: they have mechanisms for knowledge to accumulate and compound.

This looks different in every business, but some version of the same elements tends to appear. There is a consistent framework for evaluating initiatives — not just whether they succeeded or failed, but what they revealed. There is an explicit process for carrying learning from one cycle into the design of the next. There is a tolerance for initiatives that take longer than one quarter to show results, because the team understands the difference between something that is building toward an outcome and something that is simply not working.

Perhaps most importantly, there is a shared understanding that the goal of any given initiative is not just the immediate result but the reduction of uncertainty — the narrowing of the gap between what the business believes about its customers, its market, and its own capabilities, and what is actually true.

The management theorist Karl Weick described organisations as sense-making systems — entities whose primary activity is interpreting ambiguous signals from the environment and using those interpretations to guide action.⁶ By this definition, a business that resets constantly is one that has given up on sense-making. It is no longer trying to understand its environment. It is simply trying things, hoping something will work.

That is a posture, not a strategy. And it rarely produces anything that compounds.

The Question Underneath the Activity

If your business has been actively trying things for several years without establishing a clear sense of what is definitively working and why — without a body of knowledge about your customers and your market that grows more precise with each passing year — the problem is probably not the quality of the things being tried.

It is the absence of the architecture that would allow the trying to accumulate into knowing.

That architecture is not complicated. But building it requires a decision to treat learning as a first-order outcome — not a byproduct of activity, but the thing you are most deliberately pursuing.

Most businesses have not made that decision explicitly. Which is why most businesses are still, essentially, starting from the same place they were a few years ago.

References

Eric Ries, The Lean Startup: How Today's Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses, Crown Business, 2011.
Chris Argyris, Teaching Smart People How to Learn, Harvard Business Review, May–June 1991.
Amy C. Edmondson, The Fearless Organisation: Creating Psychological Safety in the Workplace for Learning, Innovation, and Growth, Wiley, 2018.
Wolfram Schultz, Predictive Reward Signal of Dopamine Neurons, Journal of Neurophysiology, 1998.
Roger L. Martin, The Opposable Mind: How Successful Leaders Win Through Integrative Thinking, Harvard Business Review Press, 2007.
Karl E. Weick, Sensemaking in Organisations, SAGE Publications, 1995.