The research on why reflection outperforms practice — and what it means for how you architect growth

The counterintuitive finding

When given a choice between practising a task and reflecting on previously accumulated practice, most people choose to keep practising. The research suggests this preference is consistently wrong.

Di Stefano, Gino, Pisano and Staats conducted ten experimental studies with 4,340 participants to establish something that runs against the instinct of almost every entrepreneur: learning from direct experience is more effective when coupled with deliberate reflection. Participants who completed an arithmetic task and then reflected on their strategies went on to perform 18% better in a second round than those who continued practising without reflection. Less time doing, more time thinking about what was done, produced measurably better subsequent performance.

The mechanism is self-efficacy — the confidence in one’s capacity to achieve a goal. Reflection builds an accurate and updated model of one’s own competence. That model is the psychological substrate for the next performance phase. Without the pause, the self-efficacy model remains calibrated to the challenges of a previous stage — and the entrepreneur enters the next stage with misaligned confidence in capabilities that may no longer match what is required.

The field confirmation came from a Wipro study embedded in the same research programme. Groups of new employees in training were assigned to control, reflection, and sharing conditions. The reflection group outperformed the control group on subsequent tasks despite having accumulated less raw practice time. The company that appeared to be slowing down by building reflection into training was actually accelerating its learning curve — because it was converting experience into expertise rather than merely accumulating experience. These are not the same thing.

The exploitation trap that momentum builds

James March’s foundational 1991 paper on exploration and exploitation identified the structural mechanism that makes deliberate pauses necessary rather than optional. Exploitation — using existing resources, knowledge, and capabilities to improve efficiency — tends to crowd out exploration because its returns are faster, more certain, and more visible. This creates a self-reinforcing trap: organisations become progressively more efficient at their current configuration while the next stage of growth requires a fundamentally different one.

This is not a management failure, it is a structural property of how organisations learn and allocate attention. A company accelerating continuously through growth stages is, by March’s logic, progressively starving exploration in favour of exploitation. The pause at a milestone is the mechanism that breaks this trap. It creates the psychological and organisational space in which exploration becomes possible — new identity, new mental models, new strategic frames. Without it, the exploitation of what works now structurally prevents the development of what will be needed next.

What expertise research establishes about cycles

Ericsson’s deliberate practice framework provides the adjacent evidence. Expert performance is not built by continuous doing. It is built by cycles of doing, assessing, correcting, and doing again. The assessment phase — the pause — is where schema reconstruction occurs, converting raw experience into transferable expertise. A company that does not build structured reflection into its milestone architecture is not building organisational expertise. It is accumulating organisational experience. Experience without reflection produces habit. Reflection converts experience into capability.

Research on hot streaks in artistic, cultural, and scientific careers adds a temporal dimension. The highest-performance bursts are not random — they follow periods of exploration and experimentation with periods of focused exploitation. The pattern implies a natural milestone architecture: explore broadly, pause to consolidate, then exploit with focus. Organisations that skip the pause go directly from exploration to more exploration, or from exploitation to more exploitation, and miss the conversion mechanism that turns varied experience into directed high performance.

What it looks like at scale

When Satya Nadella became Microsoft’s CEO in 2014, the company was executing well on a strategy that was becoming irrelevant. His first intervention was not a product decision. It was a deliberate cultural pause — a period of explicit reflection on what Microsoft’s identity actually was, using Carol Dweck’s growth mindset framework as the diagnostic tool. The commercial recovery that followed is widely attributed to this period of intentional cultural reflection rather than to any immediate strategic move. The pause preceded the pivot. The psychological reconstruction of the organisation’s identity preceded the strategic repositioning.

The US Army’s After Action Review provides the most disciplined institutional version of the same principle. Conducted immediately after every significant operational event — not at annual reviews, but at the specific moment of task completion — the AAR asks four consistent questions: what was supposed to happen, what actually happened, why was there a difference, and what will be done differently next time. Research on AAR effectiveness consistently shows that units conducting them learn faster and sustain higher performance across changing conditions than those that don’t. The pause converts operational experience into transferable learning at the moment when the experience is most cognitively accessible.

What to actually do

The milestone pause does not require a retreat or a strategy away-day. It requires a structured moment at natural transition points — after a product launch, after a funding round, after a significant hire, after a quarter closes — in which three questions are answered explicitly rather than glossed over in the momentum of what comes next. What worked, and why? What didn’t, and why? What does that mean for how we approach the next stage? The reflection group in the Wipro study outperformed the control group. The AAR units outperformed non-AAR units. The mechanism is consistent: the pause that appears to slow progress is the investment that accelerates it.

A book worth reading alongside this

Thinking in Systems by Donella Meadows provides the most rigorous available framework for understanding why deliberate pauses function as leverage points in complex adaptive systems. Her treatment of feedback loops, delays, and the counterintuitive behaviour of systems under pressure explains precisely why the pause that appears to slow a growing company is often the intervention that accelerates it at a deeper level. For any entrepreneur who wants to understand the systems logic beneath the reflection research, it is the most intellectually grounded starting point available.

Have questions about this article?

If any part of this article raised questions you want to explore further, courbot.co is built for exactly that. It is courben.co’s AI assistant, designed around the psychology of entrepreneurship. Ask it anything from this article.

This article is for educational and informational purposes only. Sources: Di Stefano, G., Gino, F., Pisano, G. & Staats, B.R. (2014), Harvard Business School Working Paper 14-093. March, J.G. (1991), Organization Science, 2(1). Ericsson, K.A., Krampe, R.T. & Tesch-Römer, C. (1993), Psychological Review, 100(3).