Good Strategy Bad Strategy: The Difference and Why It Matters
Source: https://www.penguinrandomhouse.com/books/203906/good-strategy-bad-strategy-by-richard-rumelt/ ↗
Rumelt's argument is cutting: most of what gets called "strategy" in organisations is not strategy at all — it is goals, or slogans, or lists of initiatives.
Real strategy has a kernel: a diagnosis of the situation, a guiding policy, and a coherent set of actions.
Without all three you have bad strategy, which is often worse than no strategy because it commits resources in a direction that cannot succeed.
For product direction this is the most diagnostic book on strategy in print; Rumelt's vocabulary (kernel, bad strategy, chain-link systems) survives contact with actual strategy meetings.
Read alongside Porter for the theoretical foundations, Magretta for a compressed Porter, and Rumelt for the operational critique.
Central argument
Rumelt argues that most organisational strategy is counterfeit: it consists of aspirational goals, motivational slogans, or undifferentiated lists of initiatives dressed up as direction. Real strategy, he contends, has a kernel of three interdependent elements — a diagnosis that defines the nature of the challenge, a guiding policy that channels response, and a coherent set of actions that execute that policy. Bad strategy is not merely the absence of good strategy; it actively destroys value by committing resources and attention to directions that cannot produce leverage.
Critique
Rumelt's framework is diagnostically powerful but prescriptively thin: he is far more rigorous about identifying what bad strategy looks like than about specifying how to construct a diagnosis when the situation is genuinely ambiguous or fast-moving. His examples skew toward large industrial and military cases where hindsight makes the correct diagnosis obvious, which understates the epistemic difficulty of diagnosis in real time — a tension particularly acute in digital contexts where the competitive situation shifts before a kernel can be stabilised and executed.
Why it matters for product
For a CPO, Rumelt's kernel concept is a direct test for product strategy documents: if a roadmap or OKR set cannot be traced back to an explicit diagnosis of why the current situation is the constraint it is, it is almost certainly bad strategy by Rumelt's definition — goals masquerading as direction. His notion of chain-link systems is equally sharp for product organisation: investing in discovery or delivery velocity where the binding constraint is actually go-to-market or engineering capacity produces no leverage, which explains why many product improvements fail to move business outcomes.