A New Definition of Strategy — and Why AI Makes It Urgent

From Chandler to Porter to Mintzberg, every definition of strategy has operated at a single level. The Five Levels definition operates across all of them — and identifies what remains human in the age of AI.

The word "strategy" has been used for over two thousand years, and for most of that history it meant one thing: the art of the general. Sun Tzu wrote about it as the calculus of positioning and deception. Clausewitz defined it as the use of engagements for the purpose of war. Liddell Hart described it as the art of distributing and applying military means to fulfill the ends of policy. Each was describing a specific level — the coordination of military force toward political objectives.

When the word migrated to business in the mid-twentieth century, it carried that level-specific character with it. Drucker (1954) framed strategy around a foundational question — "what is our business?" — and the responsibility of top management to answer it. Chandler (1962) defined it as the determination of long-term goals and the allocation of resources to achieve them — and showed that organizational structure follows from that determination. Ansoff (1965) formalized strategy as a decision rule for selecting which products to develop and which markets to enter. Henderson, through the Boston Consulting Group in the 1960s and 1970s, introduced the experience curve and the growth-share matrix — tools that made portfolio-level strategy visible and manageable.

Then came the two thinkers who shaped how a generation understands the word. Porter (1980) defined strategy as the broad formula for how a business competes — choosing a unique position through a distinctive set of activities. Mintzberg (1987) pushed back, arguing that strategy is not only a deliberate plan but also a pattern that emerges in a stream of decisions, a position, a perspective, and sometimes a ploy. Between them, Porter and Mintzberg framed a debate — deliberate versus emergent, positioning versus process — that has dominated the field for four decades.

Each of these definitions captures something real. None of them answers a more fundamental question.

1. What Is Strategy Across All Levels?

What is strategy as a structural phenomenon — not at one level, not from one vantage point, but across all levels?

What is the common structure beneath what a CEO does, what an educator does, what a coach does, what a parent does, what an individual does when managing their own decisions? Is there one thing that makes all of these "strategy"?

Twenty years ago, in an article published in MIT Sloan Management Review titled "Is Strategy a Bad Word?" (Mitreanu, 2006a), I argued that the frequent failure of strategy might lie in its very definition — that the word historically implies short-term victory over competitors rather than long-term vision and lasting success. The article raised the question. It did not provide a definition.

The question was not isolated. In "Looking Down on Corporate Strategy" (Mitreanu, 2006b), I argued that the concept of strategy had been diverted by historical circumstances — that the post-war infusion of military officers into business management, combined with Keynes's macro-micro divide, had replaced the quest for a fundamental theory of business with the quest for sustainable competitive advantage. The essay closed with a question: "What would corporate strategy be if the concept had emerged not two decades after a world war, but five or more?" The Five Levels definition, proposed twenty years later, is the answer. The theoretical development behind it began in 2002 and produced its first written account in a manuscript submitted to Harvard Business Review in 2003, followed by the paper "Strategy, Redefined" (Mitreanu, 2004) — the first formal presentation of the framework that would become the Ofmos Theory of Business. These were steps along a path that has now spanned more than two decades. The definition proposed here is where that path leads.

2. The Definition and the Structure Behind It

Recent work on the structural dynamics of intelligence suggests that when any intelligent agent — human or artificial — pursues successful existence in an environment of scarcity, the pursuit organizes itself into levels of increasing complexity. This is not a design choice. It is what complexity science predicts: at each level of organization, genuinely new properties emerge that cannot be reduced to the level below (Simon, 1962; Anderson, 1972).

At the most basic level, the agent manages decisions. At the next level, the agent manages tools that shape decisions. Then offerings in a shared landscape. Then an organization coordinating across multiple domains. Then a systemic economy. Five levels — each with its own unit of analysis, its own emergent phenomena, and its own dynamics.

At each level, there is a system that the strategist manages and orchestrates. And at each level, there is a formula for sustained success — a specific configuration of actions and priorities that the dynamics at that level reward. The formula is different at every level because the phenomena are different. But the strategic task is the same: find the formula, pursue it, and adapt as the system and its environment evolve.

This architecture — described in full in the Five Business Big Pictures, a framework built on the One-Need Theory of Behavior and the Ofmos Theory of Business — produces a definition:

Strategy is the adaptive, purposeful formula for success at the broadest meaningful resolution at which a system — managed and orchestrated by a human being — can be described.

Because it derives from the architecture of the Five Business Big Pictures — and because it is the first definition that operates across all levels at which strategic thinking is practiced — we refer to it as the Five Levels definition of strategy.

A clarification on terms: the Five Levels definition of strategy and the Five Business Big Pictures framework are not the same thing. The definition is a single statement about what strategy is — the adaptive, purposeful formula for success at the broadest meaningful resolution. The framework is the architecture that identifies the five specific levels at which strategy operates, the dynamics at each level, and the formula for success at each. The definition is the distillation. The framework is the full architecture.

3. What the Definition Says — and Why Each Word Matters

Adaptive — because the system and its environment change continuously. A formula that does not adjust is a plan, and plans break on contact with reality. Strategy is what survives that contact because it adjusts.

Purposeful — because strategy requires an agent who intends, who commits to a course of action, who orchestrates. Purpose distinguishes strategy from drift. A pattern that emerges without intent may be observable, but it is not strategy — it is a consequence.

Formula for success — not "formula for winning" and not "formula for competitive advantage." Success is broader. It applies at every level, including levels where there is no direct competitor — the individual managing their own decisions, the parent guiding a teenager's development, the economy pursuing dispersion and resilience. The framework identifies a specific formula at each level, and each formula describes what sustained success requires at that level of organization.

At the broadest meaningful resolution at which a system can be described — this clause sets the boundary. A system is "meaningfully described" when it is coherent enough to have a formula for success that the strategist can identify and pursue. Below that resolution, the work is executing within a formula that someone else — or a larger system — has set. Above it, the system is too diffuse for any single agent to orchestrate. The strategist's job is to find that boundary — the largest coherent system they can manage — and to operate there.

Managed and orchestrated by a human being — because strategy is not computation. It is the act of an agent who holds a complex system together across time, under uncertainty, through judgment that integrates information no algorithm fully captures — context, purpose, competing values, and the willingness to commit when the data is incomplete. AI can inform this judgment. It cannot perform it — because performing it requires being the agent whose existence is at stake. The age of AI makes this clause urgent, and the next section explains why.

4. Why This New Definition Is Not Just More Fundamental — It Is More Useful

A bold claim requires a direct defense. The claim is not merely that this definition is more fundamental than the existing ones. It is that a definition operating at this level of generality is more practical, more learnable, and more powerful than the collection of frameworks it subsumes.

4.1. The problem with separate frameworks

Today, a strategy student — whether in a business school, a corporate workshop, or a self-directed learning path — encounters a landscape of separate tools, each developed independently, each resting on different theoretical assumptions, and each applicable at a specific level. Porter's Five Forces (Porter, 1979) describes the competitive dynamics of an industry. The BCG growth-share matrix (Henderson, 1970) classifies products in a portfolio. Dynamic capabilities theory (Teece, Pisano, & Shuen, 1997) describes how organizations adapt. The balanced scorecard (Kaplan & Norton, 1992) measures organizational performance. Jobs-to-be-done (Christensen et al., 2016) identifies customer needs. Core competencies (Hamel & Prahalad, 1990) identifies organizational strengths. The value chain (Porter, 1985) maps internal operations. Disruptive innovation (Christensen, 1997) explains how entrants unseat incumbents. Blue ocean strategy (Kim & Mauborgne, 2005) describes the creation of uncontested market space.

Each of these is useful. Each captures something real. And each leaves the learner with the same problem: how do they all fit together? What connects the five forces to the growth-share matrix to dynamic capabilities to the value chain? The student must learn each one separately, understand each one's assumptions, and then do the integration work themselves — assembling a mental model from parts that were never designed to be assembled.

Most never do. The result is what business education has produced for decades: practitioners who know individual frameworks but cannot see across them, who can classify a product on the BCG matrix but cannot explain why it ended up there, who can map a value chain but cannot predict which direction the market forces will push it.

4.2. What changes with a single generating logic

The Five Business Big Pictures is built on a single foundation: two first-principles theories — the One-Need Theory of Behavior and the Ofmos Theory of Business — mapped onto a single reference system, the Ofmos Map. Every level is an expression of the same two theories. Every framework that the student would have learned separately turns out to be a partial view of dynamics that the theories describe from first principles.

The BCG matrix becomes a simplified approximation of the Ofmos Map at Product Level — its two dimensions (market growth rate and relative market share) are market-level proxies for the dynamics the map describes from first principles. The product life cycle becomes the trajectory the theory predicts for every offering under the commoditization force. The experience curve becomes the supply-side expression of the same force. Porter's Five Forces become the competitive pressures that emerge within a tofmos — an industry viewed as a behavioral pattern — when multiple vendors respond to the commoditization force through their strategic actions. The five forces are not the causes of competitive dynamics; they are what those dynamics look like when viewed from outside. Dynamic capabilities become the organizational response to commoditization — the ability to sense, seize, and reconfigure resources as markets shift. Jobs-to-be-done becomes a need at a specific level of the Hierarchical Tree of Needs. Disruptive innovation becomes leftward product innovation — decreasing complexity to capture a position the incumbent has overshot. Blue ocean strategy becomes the creation of a new ofmos at a position on the Ofmos Map where no existing tofmos sits — the vendor creates a new offering-market pair rather than competing within an existing industry.

Each framework remains useful. But the learner who understands the generating logic no longer needs to learn each one as an independent tool and then figure out the connections. The connections are already there — because all of these frameworks are describing different aspects of the same underlying dynamics. The learner encounters the BCG matrix and recognizes it as a simplified view of something they already understand. They encounter Porter's Five Forces and see the competitive dynamics the theory predicts at Company Level. The integration work that used to be the hardest part of strategic education — the part most students never complete — is done by the framework itself.

Three consequences follow — each a direct result of having one generating logic instead of many separate frameworks.

Easier to learn. One generating logic is easier to learn than eight separate frameworks, each with its own assumptions, its own vocabulary, and its own domain of application. The student learns the One-Need Theory (how individuals behave and why value erodes), the Ofmos Theory (how the individual dynamics aggregate into market and organizational dynamics), and the Ofmos Map (the reference system that applies at every level). From that foundation, every existing framework becomes recognizable — a specific case of dynamics the student already understands.

Easier to apply across contexts. Because the same reference system operates at every level, a strategist who has internalized the framework can move between levels without switching mental models. The same map that describes a product portfolio describes an AI tool portfolio describes an economy's industrial structure. The same dynamics — commoditization, innovation, portfolio management — operate at every level, with different units of analysis but the same structural logic. A coach working with a client on individual decision-making and a CEO managing a diversified company are both working on the same map.

More powerful. A strategist carrying one unified framework sees connections that a strategist carrying separate frameworks cannot see. They see that the same force driving a product's decline in the market is the same force driving a decision's declining value in an individual's life — because both are expressions of the same learning mechanism operating at different levels. They see that a company's loss of strategic focus is structurally parallel to an individual's loss of decision coherence. They see that an economy's structural fragility follows the same logic as a product portfolio's overconcentration. These connections are invisible to the strategist who learned each level's frameworks independently — because the frameworks were never designed to reveal them.

This is the practical consequence of a definition that works at every level. It is not just a more elegant description. It is a more efficient tool — one that is easier to learn, easier to apply, and more powerful in use, precisely because it operates at the level where the generating logic is visible.

5. Why AI Makes It Urgent

AI is absorbing an expanding range of cognitive tasks — analysis, pattern recognition, optimization, classification, even elements of decision-making. The decisions that AI handles best are structured, data-rich, and optimizable within defined parameters. These are real capabilities, and they are reshaping every profession.

But the strategic task — finding the formula for success at the broadest meaningful resolution you can manage, adapting it as conditions change, holding the system together across time and uncertainty — is not a structured, optimizable problem. It is the problem of agency itself: an intelligent being, operating under scarcity, choosing what to pursue and how to pursue it, across a system too complex to optimize but too consequential to leave unmanaged.

The definition above makes this explicit. Strategy is managed and orchestrated by a human being. Not because AI is incapable — that may change — but because, at this moment in history, the orchestration of complex systems across multiple levels under genuine uncertainty is the function that remains most distinctly human. It is what AI augments but does not replace. It is what educators, coaches, L&D professionals, and parents are developing in others — whether they call it strategy or not.

Every previous definition of strategy was written before AI forced the question of what remains human. This one is written with that question at the center.

And the practical stakes are immediate. If strategic thinking is the capability that AI makes more valuable — not less — then the efficiency with which that capability can be developed matters enormously. A learning system built on eight separate frameworks, each requiring independent instruction, is slower, more fragmented, and harder to retain than a learning system built on one generating logic that reveals what all eight describe. In the age of AI, the speed and depth with which a learner can develop genuine strategic thinking is not an academic question. It is a competitive advantage — for the learner, for the educator, and for the institution.

6. Strategy Is Not a Business Concept

From Sun Tzu to Chandler to Porter to Mintzberg, every definition of strategy has been anchored to a specific domain — war, the corporation, the industry, the decision-making process. Each is useful within its domain. Each remains so.

Strategy is more fundamental than any single domain. It is a structural phenomenon — one that appears wherever an intelligent agent manages a complex system under conditions of scarcity and uncertainty. As argued in 'What Would AGI Actually Need to Succeed?' on this blog, the generating logic behind the five levels is not specific to human intelligence. It is specific to intelligence itself. Any domain where multiple intelligent agents specialize, coordinate, and exchange value under conditions of scarcity would produce the same levels of complexity — and at each level, the strategic task would be the same. That domain is business, whether the participants call it that or not. And if business is natural, then so is strategy — a phenomenon that should be studied, understood, and practiced as such.

The argument that business is a natural phenomenon — not a human invention — is explored in an upcoming companion post on this blog: "Business as a Natural Phenomenon."

The full framework is available at ofmos.com/the-strategy-framework.

The foundational theories are described at ofmos.com/the-foundational-theories.

References

Anderson, P.W. (1972). "More Is Different." Science, 177(4047), 393–396.

Ansoff, H.I. (1965). Corporate Strategy: An Analytic Approach to Business Policy for Growth and Expansion. McGraw-Hill.

Chandler, A.D. (1962). Strategy and Structure: Chapters in the History of the Industrial Enterprise. MIT Press.

Christensen, C.M. (1997). The Innovator's Dilemma. Harvard Business School Press.

Christensen, C.M., Hall, T., Dillon, K. & Duncan, D.S. (2016). "Know Your Customers' 'Jobs to Be Done.'" Harvard Business Review, 94(9), 54–62.

Clausewitz, C. von (1832). Vom Kriege. Dümmlers Verlag. [English translation: On War, Princeton University Press, 1976.]

Drucker, P.F. (1954). The Practice of Management. Harper & Brothers.

Hamel, G. & Prahalad, C.K. (1990). "The Core Competence of the Corporation." Harvard Business Review, 68(3), 79–91.

Henderson, B.D. (1970). "The Product Portfolio." BCG Perspectives.

Kaplan, R.S. & Norton, D.P. (1992). "The Balanced Scorecard — Measures That Drive Performance." Harvard Business Review, 70(1), 71–79.

Kim, W.C. & Mauborgne, R. (2005). Blue Ocean Strategy. Harvard Business School Press.

Liddell Hart, B.H. (1967). Strategy. 2nd revised edition. Faber & Faber. [First published 1954 as Strategy: The Indirect Approach.]

Mintzberg, H. (1987). "The Strategy Concept I: Five Ps for Strategy." California Management Review, 30(1), 11–24.

Mitreanu, C. (2004). "Strategy, Redefined." RedefiningStrategy™.

Mitreanu, C. (2006a). "Is Strategy a Bad Word?" MIT Sloan Management Review, 47(2), 96.

Mitreanu, C. (2006b). "Looking Down on Corporate Strategy." RedefiningStrategy™.

Porter, M.E. (1979). "How Competitive Forces Shape Strategy." Harvard Business Review, 57(2), 137–145.

Porter, M.E. (1980). Competitive Strategy: Techniques for Analyzing Industries and Competitors. Free Press.

Porter, M.E. (1985). Competitive Advantage: Creating and Sustaining Superior Performance. Free Press.

Simon, H.A. (1962). "The Architecture of Complexity." Proceedings of the American Philosophical Society, 106(6), 467–482.

Sun Tzu (c. 5th century BC). The Art of War. [Griffith, S.B. (trans.), Oxford University Press, 1963.]

Teece, D.J., Pisano, G. & Shuen, A. (1997). "Dynamic Capabilities and Strategic Management." Strategic Management Journal, 18(7), 509–533.

Cristian Mitreanu is a behavior and strategy researcher, product professional, and educator based in San Francisco. He is the founder of Ofmos Universe — The Human Strategist Platform™.