Good Strategy/Bad Strategy

Strategy is not the ambition on the first slide

Good Strategy/Bad Strategy begins with a provocation: much of what organizations call strategy is not strategy. It is a performance goal, a list of values, a vision of success, or a collection of initiatives that no one has chosen among. Richard Rumelt does not object to ambition. He objects to using ambition as a substitute for confronting a difficulty and deciding how to overcome it.¹

The book's examples range among corporations, governments, schools, wars, and technologies. The range supports a general claim: strategy is a response to a challenge in which thought identifies leverage and action concentrates it. A good strategy need not predict the whole future or specify every operating step. It gives the organization a coherent way to move through the obstacle that matters now.²

Bad strategy avoids the hard choice

Rumelt's contemporary summary of bad strategy identifies several recurring symptoms. Fluff restates ordinary ideas in inflated language. A failure to face the challenge leaves the real obstacle unnamed. Mistaking goals for strategy declares a desired result without explaining what will make it possible. Bad strategic objectives are either a disorganized list or targets that cannot be reached with available power.³

These failures are not simply errors of writing. They often protect coalitions. Naming the central challenge can expose a failing business model, cherished program, political conflict, or missing capability. Concentrating resources creates losers among internal priorities. A broad vision lets everyone hear their project affirmed; an actual strategy says what will not be done.⁴

The critique is useful because it turns a document review into organizational diagnosis. If every objective has top priority, authority has not made a choice. If the account of the environment contains no obstacle, the plan cannot explain why ordinary competent action is insufficient. If a numeric target is not tied to a causal mechanism, it is pressure, not guidance.⁵

The kernel connects interpretation to action

Rumelt's positive structure has three parts. A diagnosis simplifies the complex situation by explaining the nature of the challenge. It names a pattern that distinguishes essential from secondary facts. A guiding policy sets an overall approach for dealing with the diagnosed obstacle. It channels action without trying to predetermine every local decision. Coherent actions are coordinated commitments, resource allocations, and moves that enact the policy and reinforce one another.⁶

The kernel is memorable because it is small, but it is not easy. A diagnosis is an inference open to disconfirmation, not a problem label copied from a workshop. A guiding policy must exploit an advantage, asymmetry, or source of leverage. Coherence requires choices across functions whose local incentives differ. The actions also create evidence: if they do not change the predicted constraint, the diagnosis or policy may be wrong.⁷

Rumelt's sources of strategic power include leverage, proximate objectives, chain-link systems, design, focus, growth from advantage, and exploiting waves of change. In a chain-link system, performance is limited by its weakest link; improving already strong links can waste resources. A proximate objective turns an ambiguous distant aspiration into something the organization can presently achieve. Design coordinates choices so the whole produces an effect no isolated move could.⁸

Focus creates moral as well as operational consequences

The book is strongest on intellectual honesty and concentrated action. Its cases are illustrative rather than a controlled empirical test of the kernel. A story's coherence after the outcome can make successful action look more designed than it was; failed strategies and emergent adaptations are harder to see. Rumelt's argument and the independent evidence needed to establish each case remain distinct.⁹

More importantly, a good strategy in Rumelt's technical sense can serve an unjust purpose. A diagnosis can treat workers, competitors, or a community as the obstacle. Leverage can exploit dependency. Coherence can make harm more effective. Ethical legitimacy must therefore be an additional test, not an assumption hidden inside “good.”¹⁰

Structured reading paths distinguish contrast from influence

The dependency on Competitive Strategy is an editorial contrast: Michael E. Porter analyzes industry structure and position, while Rumelt organizes strategy around a diagnosed challenge. The dependency on The Rise and Fall of Strategic Planning places Henry Mintzberg's critique of formal planning beside Rumelt's attempt to recover strategic judgment. These comparisons do not establish a direct debate, common method, or historical influence.¹¹

Paths to strategy, competition, and adaptation, executive attention, information, and organizational sensing, and organizational intelligence connect diagnosis and focus to broader questions of adaptation and selective attention. Benefit for all life marks the additional legitimacy test that Rumelt's technical kernel does not supply. No structured impact observations or typed relations are encoded, so the reading paths support neither effect-size estimates nor claims that the linked works share one definition of strategy.

Use Rumelt by writing the diagnosis as a falsifiable causal claim, the guiding policy as a boundary on choices, and the actions as a mutually reinforcing set. Add a fourth question the kernel leaves open: if this works exactly as intended, which lives gain agency, which lose it, and what evidence could make the organization stop? The selected sources are the book record, its publisher, and Rumelt's own essay; they contain no representative worker, competitor, community, or ecological outcome evidence.¹²