Richard Rumelt argues that good strategy is not a vision statement, a set of ambitions or a long list of priorities. It begins with an honest diagnosis of the pivotal challenge, establishes a guiding policy for addressing it and translates that policy into coherent action.
The book’s distinctive contribution is its insistence on choice and concentration. Rumelt shows how organizations mistake goals, slogans and financial targets for strategy, while avoiding the difficult work of identifying leverage. His kernel of diagnosis, guiding policy and coherent actions offers leaders a practical test for whether a strategy can actually direct decisions.
For executives, the implication is to reduce complexity before adding initiatives: name the crux, decide what will not be pursued and align resources behind a small number of mutually reinforcing actions. The framework is especially useful when priorities have accumulated faster than capacity, although diagnosis still requires judgment and cannot be reduced to a template.
OutcomesLab profiles Good Strategy/Bad Strategy because it provides one of the clearest foundations for strategic coherence. It explains why focus creates leverage, why competing priorities produce Dilution Drag and why effective strategy must connect a precise problem to coordinated action.
How the argument works
The argument works by defining strategy as a coherent response to a consequential challenge, built from diagnosis, guiding policy and coordinated action. Rumelt’s kernel begins with diagnosis because leaders cannot choose a useful response until they explain what is actually preventing progress.
A good diagnosis simplifies reality without trivializing it. It identifies the pivotal structure of the challenge—such as a bottleneck, competitive asymmetry, capability gap or reinforcing dynamic—and separates it from symptoms. The diagnosis is a judgment about causality, not a longer description of the situation.
The guiding policy sets the overall approach. It channels action without prescribing every move and creates boundaries that allow decentralized decisions to remain consistent. A policy has force when it exploits leverage, concentrates advantage or coordinates responses that would otherwise conflict.
Coherent actions then make the policy real. Their power comes from interaction: one action creates conditions that increase the effect of another. Proximate objectives can convert an ambiguous challenge into achievable steps, while design coordinates resources against a source of leverage.
Bad strategy interrupts this chain. Fluff disguises weak thinking, failure to face the challenge removes the basis for choice, mistaking goals for strategy substitutes desire for method, and too many objectives scatter resources.
The causal mechanism is coherence under constraint: diagnosis focuses attention, guiding policy directs choices, coordinated actions combine resources and the resulting concentration creates leverage. Strategy is effective because the parts reinforce one another against a problem that matters, not because the organization has expressed a more ambitious destination.
The kernel is therefore both an analytical structure and a test of whether action has been deliberately designed.
What the book gets right
The book’s enduring contribution is drawing a hard boundary between strategy and the language organizations commonly use in its place. Ambitions, financial targets, values and long priority lists may all be important, but none explains how a challenge will be overcome.
The kernel is memorable because each element forces a different kind of reasoning. Diagnosis asks what is true; guiding policy asks how the organization will respond; coherent action asks what will actually move. Weakness in one element cannot be repaired by eloquence in another.
Rumelt’s emphasis on leverage is equally important. Strategy is not proportional allocation across stakeholder requests. It combines resources where an asymmetry, constraint or sequence can produce disproportionate effect.
The critique of bad strategy is also operationally valuable. Fluff and goal proliferation are not merely communication flaws; they hide the absence of trade-offs and allow resources to remain dispersed.
Most importantly, the book treats coherence as a design property. Actions do not become strategic because they support the same headline. They become strategic when their interaction makes the chosen response more powerful. That is a demanding but useful test for any enterprise strategy.
It gives boards and executive teams a concise challenge: if they cannot state the diagnosis and guiding policy, they should not confuse an approved plan with an agreed strategy.
Where the argument has limits
The kernel clarifies the structure of strategy but does not guarantee the quality of the diagnosis on which everything depends. A confident executive team can create a coherent response to the wrong problem, especially when power or prior commitments shape which evidence is admitted.
The examples are persuasive but often retrospective. Once an outcome is known, the pivotal challenge and source of leverage can appear clearer than they were at the time. Leaders need explicit competing diagnoses and disconfirming evidence to reduce narrative certainty.
Guiding policies can also remain too broad. “Focus on the core” or “use our data advantage” may sound directional while leaving resource and capability choices unresolved. The policy must discriminate among plausible actions.
The framework gives less attention to adaptation after commitment. In uncertain environments, leaders need a method for preserving learning and changing the diagnosis without turning every setback into a strategic reset.
Implementation politics are also underplayed. Coherent action frequently requires stopping work, moving authority and disappointing stakeholders. A strong kernel can still fail against Resource Gravity.
Use the framework with challenge, evidence and governance. State what would make the diagnosis wrong, what the policy excludes and which resources will move. Coherence is a test of internal logic, not proof that the strategy will succeed.
How it connects to Strategic Coherence
Good Strategy/Bad Strategy provides the clearest conceptual kernel for Strategic Coherence: a precise challenge, a guiding response and mutually reinforcing action.
The strongest connection is the Coherence Premium. Actions create more value together when they express the same causal logic. A budget change, capability investment and operating decision should each strengthen the guiding policy rather than merely share a strategic label.
The book also grounds the Focus Multiplier. Focus begins with diagnosis: concentration is valuable when directed at the pivotal challenge and source of leverage. Fewer priorities without better diagnosis can simply concentrate resources on the wrong problem.
A third connection is Dilution Drag. Long lists of objectives avoid comparison and allow stakeholders to retain their claims. The cost appears through divided resources, slower decisions and weak cumulative effect.
The tension is between coherent commitment and Optionality. The kernel requires a clear current response, while uncertainty may justify experiments. Leaders should distinguish actions that execute the policy from tests designed to update its assumptions, with different funding and review rules.
What Rumelt adds to OutcomesLab is the sequence from diagnosis to coherence. Alignment should not begin around a slogan; it should begin around an explanation of the challenge. OutcomesLab adds back the organizational test: budgets, capabilities, measures and stopped work must reflect the kernel. Otherwise Resource Gravity will preserve the old strategy beneath the new language.
The kernel also improves Decision Velocity. A discriminating guiding policy allows teams to resolve secondary choices without repeatedly escalating for interpretation, provided its boundaries and assumptions are understood across the organization.
Put it to work
Use the kernel when a strategy is dominated by ambitions, themes and objectives but cannot explain the pivotal challenge or how coordinated action will overcome it.
Build and test the kernel:
- What is the most consequential obstacle, and what evidence supports that diagnosis?
- Which competing diagnosis could also explain the facts?
- What guiding policy directs action and excludes plausible alternatives?
- Which actions reinforce one another against the diagnosed challenge?
- What result or evidence would require the kernel to be revised?
Write the diagnosis before the goals. Invite a team to argue for an alternative causal explanation and identify evidence capable of separating the two. Make the guiding policy specific enough to resolve a real trade-off.
Map current initiatives against the kernel. Stop or redesign work that does not support the policy, and move resources toward the coherent set. Define proximate objectives where the final outcome is too distant to guide immediate action.
The common misapplication is completing the three boxes without facing political consequence. Publish what will not be pursued, assign authority and review evidence at a predetermined point. A kernel becomes strategy only when it changes the organization’s pattern of action.
Retain the rejected diagnosis and evidence so future review can distinguish genuine learning from a change in executive preference.
