Strategy is one of the most used and least understood words in business. Executives describe their strategy constantly — in board presentations, investor calls, all-hands meetings, press releases. Most of what they’re describing is not strategy. It’s goals, vision, aspirations, or operational priorities dressed up in strategic language.
What Bad Strategy Actually Is
Rumelt begins with what he calls “bad strategy” — not strategy that was well-executed but chose the wrong direction, but documents and approaches that are not strategy at all, regardless of the label slapped on them.
Bad strategy has four recognizable hallmarks:
Fluff: High-sounding words that restate the obvious or assert something false. “Our strategy is to be the world’s most customer-centric company” is not a strategy. It’s a value statement pretending to be one. It says nothing about what decisions have been made or what trade-offs have been accepted.
Failure to face the challenge: A strategy that doesn’t clearly define the obstacle it’s designed to overcome. If it doesn’t begin with a clear-eyed analysis of the specific challenge the organization faces, it can’t possibly describe an adequate response to that challenge. Plenty of strategy documents avoid naming the challenge, because naming it honestly would be uncomfortable.
Mistaking goals for strategy: Stating targets — grow revenue 20%, increase customer satisfaction, expand to new markets — as if they were strategy. Goals describe where you want to go. Strategy describes the approach that gets you there. Different things. A plan to “grow revenue 20%” is not a strategy — it’s an aspiration. The strategy is the specific approach to the specific market with specific competitive advantages that makes 20% growth achievable.
Bad strategic objectives: Vague objectives that don’t connect to real actions, or objectives that are unreachable or irrelevant to the actual challenge. “Deliver world-class customer service” is a bad strategic objective. Unmeasurable, unspecific about what will change, no guidance for whoever has to execute it.
“Bad strategy is not simply the absence of good strategy. It grows out of specific misconceptions and bad habits, and it can poison a whole organization.”
The Framework: The Kernel of Good Strategy
Good strategy, Rumelt argues, always has the same basic structure — what he calls “the kernel.” Three parts:
- Diagnosis: A clear definition of the challenge. What, specifically, is the problem or opportunity? What are the constraints and resources? The diagnosis should simplify the often overwhelming complexity of an actual organizational situation by identifying certain aspects as critical. A good diagnosis names the problem honestly, including the uncomfortable parts.
- Guiding Policy: An overall approach for dealing with the obstacles identified in the diagnosis. The guiding policy channels action in certain directions without defining exactly what shall be done. It rules out a wide array of actions and focuses resources on a narrower set of coherent moves. This is the key: a good guiding policy excludes options, which is why most organizations resist it.
- Coherent Actions: Coordinated policies, resource commitments, and practical steps that carry out the guiding policy. The actions must reinforce each other — not simply be a list of independent initiatives that happen to share a slide deck.
The diagnosis-policy-action structure sounds simple. It is. But it gets violated constantly in real organizational life, because each element requires something most organizations are bad at: honesty (diagnosis), focus (guiding policy), and coordination (coherent actions).
The diagnosis step fails most often because naming the real problem means admitting uncomfortable truths. A company losing market share to a cheaper competitor needs to diagnose that competition honestly. But doing so means acknowledging either that the product isn’t worth the premium or that the go-to-market approach is wrong — both of which implicate people still in the room. The result is a diagnosis that blames external factors (macroeconomic conditions, customer preference shifts) rather than internal ones, and a strategy built on a false foundation.
The guiding policy step fails because it requires actually choosing. An organization trying to be everything to everyone — premium and value, innovative and reliable, global and locally customized — has no guiding policy. It has competing priorities. Real strategy means saying no to some good options to focus on the best ones. Every board meeting where every proposed initiative gets funded is a strategy failure masquerading as decisive action.
The coherent action step fails because most organizations are structured in ways that make coordination difficult. Different departments optimize for different goals. Incentives don’t align. Budget processes favor independent projects over integrated ones. The result: a strategy whose elements don’t reinforce each other, and a total effect less than the sum of the parts.
use and the Source of Strategic Power
One of the most important concepts in the book is use. Good strategy doesn’t try to do everything at once — it identifies the pivot point where focused effort produces disproportionate results. Rumelt’s favorite image is a fulcrum: a small force applied in the right place can move an enormous weight. Strategic use is finding the organizational or competitive fulcrum where resources produce asymmetric outcomes.
Which is why resource advantage alone is not strategy. A company with more money, more people, more market presence than its competitor still needs to identify the specific high-use points where those resources should land. Spreading resources evenly across all opportunities is the strategic equivalent of pushing on a wall everywhere at once — the distributed force rarely produces concentrated results.
The search for use requires asking: where in this competitive or organizational landscape would a concentrated application of specific advantages produce the most significant outcome? Not always obvious, and it requires the diagnostic work to find it. But when it’s found, it produces the characteristic signature of good strategy — achieving more with less, not by magic but by concentration.
Historical Case Studies: Where Strategy Actually Came From
Rumelt’s most effective teaching method is historical case study. His examples range from military history (the Desert Storm campaign, the Battle of Marathon) to corporate history (Apple’s return from near-bankruptcy, Walmart’s early competitive strategy, the containerization of shipping), and they’re consistently more illuminating than abstract principle.
The Apple example is particularly instructive. When Steve Jobs returned to Apple in 1997, the company had twelve computer product lines and was months from bankruptcy. Jobs’s strategy wasn’t improving all twelve product lines or developing a comprehensive new vision for the company. His diagnosis was that Apple had lost coherence — trying to be too many things, having abandoned the design-focused, user-experience-centered identity that had differentiated it. His guiding policy was returning to that core identity and ruthlessly eliminating everything that didn’t serve it. His coherent actions included cutting the product line from twelve to four, canceling projects, and concentrating resources on a small number of products that could exemplify the Apple experience at its best.
The iMac wasn’t the product of a visionary inspiration. It was the result of a strategic decision to concentrate all of Apple’s industrial design capability on a single consumer computer. Focus produced a product meaningfully better than anything Apple had made in years, because it had the full attention of the people capable of making it excellent.
That’s what good strategy looks like from the inside: not ten initiatives in parallel, but one concentrated push in the highest-use direction.
“A strategy is like a lever that magnifies force. The secret of strategy is finding the fulcrum where the lever will work.”
The Competitive Advantage Question
Rumelt’s treatment of competitive advantage is more rigorous than most strategy writing. His central point: sources of advantage are specific, not generic. “We have a strong brand” is not a source of competitive advantage until it’s specified what the brand actually enables that competitors can’t match, in which specific markets, against which specific competitors.
Sustainable competitive advantage, in Rumelt’s framework, comes from chains of advantage — multiple reinforcing elements individually copyable but collectively difficult to replicate. Walmart’s early advantage wasn’t just logistics or pricing — it was the specific combination of logistics, pricing, store format, supplier relationships, and technology investment reinforcing each other. A competitor could copy any one element without threatening the whole, because the advantage lived in the system, not the components.
Direct implication for how you think about your own competitive position. If the advantage is a single feature or capability, it’s more fragile than it appears — a well-funded competitor can usually match a single feature. If the advantage is a system of reinforcing capabilities, it’s more durable, because copying the system requires replicating all the elements simultaneously — much harder than matching any one of them.
Building a chain of advantage requires patience and a willingness to invest in capabilities that don’t produce immediate returns. Most organizational pressure runs the other direction — toward activities producing near-term measurable results rather than long-term structural advantage. Strategy requires resisting that pressure enough to invest in the capabilities that will matter in three to five years, not just the quarter.
Why Strategy Is Uncomfortable
The most honest and most useful section of the book deals with why good strategy is so rare even among intelligent, experienced executives. Rumelt’s answer: because the fundamental act of strategy is uncomfortable. It requires:
Honest diagnosis of failure or weakness. Most organizations carry a history of decisions that didn’t work, strategies that didn’t succeed, investments that didn’t pay off. Naming these honestly is the beginning of good diagnosis. But doing so implicates people still in power, opens old wounds, challenges organizational self-image. The default is papering over the diagnosis with optimistic framing.
Saying no to good ideas. Every budget season in every organization produces more good ideas than resources to pursue them. Good strategy requires choosing among good options — actively declining to pursue some of them. That creates disappointed stakeholders, political friction, and the risk of being wrong — declining the option that turns out to be the best one. The comfort of “we’ll try everything” avoids that risk at the cost of actually doing anything well.
Accepting uncertainty about outcomes. A real strategic bet accepts the possibility of being wrong. Executives who’ve spent their careers avoiding the appearance of failure are structurally resistant to making real bets. The result is strategies vague enough to declare success regardless of outcomes — which is to say, not strategies at all.
The political economy of organizations pushes toward bad strategy. Easier to get consensus around goals than choices. Easier to describe aspiration than diagnosis. Less risky to distribute resources than to concentrate them. Every pressure in the typical organizational environment runs against the specific things good strategy requires. The executives who produce good strategy have usually found ways to insulate the strategic process from those pressures, at least long enough to make real choices.
Personal Strategy: The Same Framework
Rumelt focuses on corporate and military strategy, but the kernel framework applies with equal force to personal strategic decisions: career pivots, major investments of time or resources, decisions about which skills to develop. Most people’s “personal strategy” is a list of goals — be healthier, make more money, develop leadership skills — rather than a diagnosis-policy-action structure. Same result as at the organizational level: goals that don’t get achieved because the approach to achieving them was never specified.
A personal application of the kernel: the diagnosis is an honest assessment of the specific gap between where you are and where you want to be, including the uncomfortable parts — what’s actually holding you back, not what sounds respectable to admit. The guiding policy is the chosen approach for closing that gap, which necessarily rules out other approaches. The coherent actions are the specific changes in how time and resources get spent that execute the guiding policy.
Most people skip the diagnosis (too uncomfortable) and the guiding policy (too limiting) and jump straight to actions that don’t cohere with each other because they’re not based on anything. Books get read, courses get attended, productivity systems get tried, habits get built that don’t connect to any underlying strategic logic. The efforts aren’t wrong — they’re just not aimed at anything specific.
This connects to our exploration of self-discipline as the foundation of resilience — the relationship between focused effort and meaningful results. For the leadership dimension of strategic thinking, see our piece on authentic leadership. And for how concentration of effort connects to personal performance, our guide to focus and concentration is directly relevant.
“The core of strategy work is always the same: discovering the critical factors in a situation and designing a way of coordinating and focusing actions to deal with those factors.”
The Denial of Proxies: When Metrics Replace Judgment
One of the most insightful and practically useful sections of Good Strategy Bad Strategy addresses what Rumelt calls “the denial of proxies” — the organizational tendency to replace real goals with measurable proxies for those goals, then optimize the proxy rather than the goal. So common that it’s become a structural feature of most large organizations, and understanding it explains why strategy consistently degrades even in organizations staffed with smart, motivated people.
The classic example is teaching to the test. The real goal is student learning — genuine understanding and capability that serves students throughout their lives. The proxy is test performance — measurable, comparable, easily aggregated. When test performance becomes the target rather than the proxy, teachers optimize for test performance rather than learning. Students who can ace the test without genuinely understanding the material become indistinguishable, within the measurement system, from students who’ve genuinely learned. The metric has become a substitute for the thing it was supposed to measure.
This exact dynamic shows up in every domain where complex goals meet institutional pressure for simple measurement. Customer satisfaction scores replacing actual customer satisfaction. Employee engagement surveys replacing actual employee engagement. Quarterly revenue metrics replacing genuine business value creation. Safety incident counts replacing actual safety culture. In each case, the proxy starts as a useful indicator and becomes a target the actual goal gets sacrificed to achieve.
The strategy implication: any strategic objective that can be easily gamed will eventually be gamed, usually by the most capable people in the organization — the ones with the most incentive and ability to optimize for metrics. Good strategy requires building in mechanisms to assess whether proxy achievement is tracking real goal achievement, and a willingness to update the measurement once the proxy has decoupled from the goal it was supposed to represent.
Strategy in Practice: The Diagnosis That Nobody Wants to Make
Rumelt’s most practically valuable contribution isn’t a framework or a set of principles but a description of a specific organizational dynamic: the systematic avoidance of honest diagnosis. Not random — it follows predictable patterns Rumelt identifies with precision, and understanding them helps recognize when your own organization is engaged in it.
The first pattern is the reframe into aspiration. When the real diagnosis is uncomfortable (“we’re losing because our product is genuinely inferior to the competitor’s on the dimensions customers care about”), the organization reframes it into aspiration (“we need to build a world-class product that exceeds customer expectations”). The aspiration feels like diagnosis because it’s in response to the problem. It isn’t — it describes a future state rather than explaining the current one, and provides no actionable guidance about what specifically needs to change.
The second pattern is external attribution. (“Our results are disappointing because the market is soft.” “Customer acquisition costs are high because the digital advertising environment has become more competitive.” “Retention is down because customers have more choices than they used to.”) These statements may be true. They’re also deflections. They describe external conditions rather than internal responses to those conditions. A diagnosis attributing the problem entirely to external factors cannot produce an actionable internal response.
The third pattern is the complexity shield. (“It’s a multifaceted problem that requires a comprehensive response across multiple dimensions.”) Technically true of almost every significant organizational problem. Also a way of avoiding the diagnostic work that would reveal which of those dimensions is most important and most actionable. Complexity is real, but “it’s complicated” is not a diagnosis.
Recognizing these patterns in an organization’s own strategic discussions is the beginning of doing the diagnostic work that produces real strategy. The question that cuts through all three: “If we had to bet everything on one change that would most improve our competitive position, what would it be?” That question forces the specificity the patterns are designed to avoid.
Why Most Strategic Plans Fail Before They Start
Rumelt is explicit about something most strategy consultants and business school professors aren’t: most strategic plans fail not in execution but in the planning process itself. They fail because the planning process is optimized for producing a document rather than making real choices, and the document produced provides cover for everyone involved while committing no one to anything specific.
The typical strategic planning process: each department or business unit develops a list of priorities and resource requests. These get compiled into a company-wide “strategic plan” that’s essentially a negotiated political document — reflecting what each stakeholder group needed to see in it to endorse it, not what would actually improve the company’s competitive position. The plan gets presented to the board, approved, circulated internally, and filed. Execution happens according to existing operational priorities rather than the plan, because the plan was never specific enough to actually guide decisions.
Rumelt’s alternative: separate the strategic planning process from the operational planning process, and run the strategic process explicitly as a choice-making exercise rather than a priority-listing exercise. The output should be a short list of genuine choices — options evaluated and rejected, and a smaller set selected and given concentrated resources. If the “strategy” doesn’t include a list of what’s been decided NOT to do, it’s probably not a strategy.
The Problem With “Best Practices”
Rumelt has a sharp critique of the “best practices” culture dominating management consulting and business education. The idea is straightforward: if something works well in one company, study what they’re doing and apply it elsewhere. The problem, which Rumelt articulates clearly, is that best practices are almost always derived from environments that differ from the one they’re being applied to, in ways that matter for whether the practice will actually produce the same results.
A practice that’s “best” in a specific competitive context — a specific market structure, a specific customer segment, a specific technology environment — may be merely adequate or actively harmful in a different context. The organizations that achieved superior results with the practice typically had other supporting conditions (talent, culture, relationships, market timing) not captured in the description of the practice itself. What looked like the source of their success was actually one component of a system, and transplanting the component into a different system doesn’t transplant the results.
Related to Rumelt’s broader point about diagnosis: before adopting any strategic approach, the specific challenge the organization faces, the specific resources and constraints in play, and whether the approach under consideration actually suits the specific situation all need to be understood. The diagnosis step is what makes best practices useful rather than cargo-cult behavior — applying the form of what successful organizations do without understanding the substance of why it worked for them.
Building a Strategy That Survives Contact with Reality
The final practical contribution of the book is Rumelt’s guidance on what makes a strategy durable — able to survive the inevitable changes in competitive environment, technology, and customer preference that occur between when a strategy is set and when it’s fully executed.
The core insight: durable strategies are built on advantages that are structural rather than situational. Situational advantages — being first to a market, holding a specific technology lead before competitors catch up, benefiting from a regulatory environment that may change — are real but temporary. They should be exploited aggressively while they last, and used to fund the development of more structural advantages. Betting the strategy entirely on situational advantages is building on sand.
Structural advantages — capabilities that are expensive and slow to develop, that compound over time through learning and investment, that are embedded in culture and process rather than in any single product or person — are more durable because they’re harder to replicate. Also harder to build, which is precisely why they’re worth the investment. The organizations still performing well a decade from now are almost always the ones that invested in structural advantages while their competitors were harvesting situational ones.
The practical question to ask about any strategic position: what would it take for a well-resourced competitor to replicate this within two years? If the answer is “not much” — buy the technology, hire the people, copy the approach — the advantage is situational and fragile. If the answer is “they’d need to rebuild the customer relationships built over ten years, develop the proprietary data accumulated, and replicate the operational culture that produces our consistency” — that’s beginning to look structural.
Strategy at the Personal Scale
Rumelt focuses exclusively on organizational and competitive strategy, but the kernel framework — diagnosis, guiding policy, coherent actions — applies to personal strategic decisions with the same clarity. Worth explicitly applying the framework to one’s own life, because most people’s “personal strategy” is functionally indistinguishable from what Rumelt describes as bad organizational strategy: aspirations stated as strategy, goals listed without coherent approaches for achieving them, efforts distributed across too many priorities to produce excellence in any of them.
A genuine personal strategic diagnosis starts with the honest question: what is specifically holding you back from the outcomes you want? Not “I need to improve generally” or “I should focus more.” The specific gap between current reality and desired outcome, stated as concretely as possible. That diagnosis might reveal that the real constraint is a specific skill gap, a specific relationship that needs building, a specific resource allocation problem, or a specific structural obstacle in your environment. Whatever it is, naming it honestly is the prerequisite for addressing it.
The guiding policy then asks: what single approach addresses this specific constraint most directly? And what does choosing that approach mean isn’t being pursued? The constraint most people’s personal strategy fails at is this one: they want to pursue everything simultaneously and treat the resulting diffusion of effort as acceptable. Rumelt’s framework forces the choice explicitly. Developing deep expertise in a field means not pursuing the adjacent opportunities that look attractive but would diffuse focus. Investing seriously in the relationships that matter most to long-term goals means not investing equally in all the relationships that exist. The policy has to exclude some options to give any options a real chance of working.
The coherent actions are then the specific changes in how time, money, and attention get allocated that execute the guiding policy. Not a list of habits or goals — a specific set of changes to current behavior that directly serve the identified policy. The coherence test: do all the actions point in the same direction and reinforce each other? Or is the list a collection of independent improvements that happen to share a slide? If the latter, it’s not strategy. It’s a self-improvement wish list.
Inertia as Opportunity
Rumelt identifies inertia — the tendency of established organizations to resist change even when change is clearly needed — as one of the most reliable and underutilized sources of competitive opportunity. Every incumbent in every industry is constrained by the organizational, process, and cultural investments it’s made in its current approach. These constraints are invisible to the incumbent, because they feel like strengths rather than liabilities, but they’re both at once. Strengths in the current environment, liabilities in any significantly different one.
The strategic opportunity: correctly identify the specific inertia constraining the leading competitors in a market, and design an approach that exploits the specific gap that inertia creates. More specific than “be more innovative than incumbents.” It requires understanding exactly which changes the incumbent cannot make quickly, why (structural commitments, customer relationships, organizational identity, incentive structures), and where the gap between their current position and an improving market creates a window. That window is typically narrower than challengers assume, and narrower than it looks in retrospect.
The Verdict on Good Strategy Bad Strategy
The most intellectually honest strategy book in print. Rumelt doesn’t sell hope or systems or frameworks that will transform your organization. He describes what strategy actually is — a specific intellectual discipline requiring honest diagnosis, focused choices, and coherent action — and why the conditions that produce it are rare and worth protecting when found.
The book is most valuable for people who’ve spent time in organizations producing the consulting-deck version of strategy — slide packages full of pillars and priorities and values that everyone applauds and no one follows. Reading Rumelt supplies the vocabulary to name what’s missing and the framework to identify what a real alternative would look like. It won’t make the organizational politics easier. But it makes the intellectual problem clearer, which is the prerequisite for anything else.
Read it alongside Competing Against Luck (Christensen) and Good to Great (Collins) for a comprehensive treatment of competitive positioning. Use it in conversation with the next strategy cycle — not to impose a framework but to ask the diagnostic questions the kernel demands. What is the specific challenge? What approach has been chosen, and what does it rule out? Are the actions actually coherent with each other?
If those questions feel uncomfortable, they’re probably being asked in the right place. For the mental discipline required to sustain rigorous strategic thinking over time, our piece on mental toughness is directly relevant. Rumelt’s approach to making difficult choices under uncertainty also connects to our exploration of grit and resilience — specifically the sections on maintaining direction when the path is unclear.
What People Ask About Good Strategy Bad
What is the difference between good strategy and bad strategy?
Good strategy identifies a specific challenge, chooses a focused approach to address it, and coordinates coherent actions that execute the approach. Bad strategy substitutes goals, values, and aspirations for actual choices, uses vague language that sounds strategic without specifying anything, and avoids the honest diagnosis of what’s actually wrong.
What is the “kernel” of strategy?
Rumelt’s term for the basic structure present in every good strategy: diagnosis (what’s the actual challenge?), guiding policy (what’s the overall approach to that challenge?), and coherent actions (what specific steps execute the policy?). All three elements are necessary. Missing any one produces something that isn’t actually strategy.
Why do most organizations produce bad strategy?
Because good strategy requires uncomfortable things: honest acknowledgment of problems and weaknesses, genuine choices that exclude good options, and acceptance of uncertainty about outcomes. The organizational pressures that typically dominate — consensus-building, face-saving, distributing credit — all run against these requirements.
Is Good Strategy Bad Strategy useful for small businesses?
Very much so. The scale of examples varies, but the framework is scale-independent. The diagnostic questions (what’s the actual challenge?), the policy questions (what approach have we chosen and what does it exclude?), and the coherence questions (are our actions reinforcing each other?) apply to a ten-person company as directly as to a corporation.
What is strategic use?
Finding the specific point in a competitive or organizational situation where concentrated effort produces disproportionate results. Good strategy identifies use points and concentrates resources there rather than spreading effort evenly across all opportunities.
How does Rumelt define competitive advantage?
As specific capabilities or positions that enable you to do something competitors cannot match in particular markets. He emphasizes that sustainable advantage comes from chains of reinforcing capabilities that are collectively difficult to replicate even if individual elements can be copied.
What are Rumelt’s four hallmarks of bad strategy?
Fluff (high-sounding language that says nothing specific), failure to face the challenge (avoiding honest diagnosis), mistaking goals for strategy (declaring targets rather than choosing approaches), and bad strategic objectives (vague, unmeasurable, or unconnected to real action).
How does Rumelt use historical examples in the book?
As the primary teaching method. He draws on military history (Desert Storm, Marathon), corporate history (Apple, Walmart, Intel), and geopolitical cases to illustrate the principles. The historical cases are typically more illuminating than the abstract principles because you can see the full context of the decision and its consequences.
What is “inertia” in Rumelt’s framework?
One of the most reliable sources of competitive opportunity. Established organizations resist change even when change is clearly needed — because of routines, incentives, and political structures that protect existing positions. Competitors who correctly diagnose this inertia can build advantages faster than the incumbent can respond.
How does Good Strategy Bad Strategy relate to other strategy books?
It’s more intellectually rigorous than most popular strategy books (Blue Ocean Strategy, Playing to Win) and more practically grounded than academic strategy texts. It’s best read alongside competitive strategy literature (Porter) and organizational learning literature (Senge) for the full picture. Rumelt is stronger on the cognitive and analytical elements of strategy than on the execution and culture elements.
Related: The Art of Happiness Summary
Related: Discipline Is Destiny Summary: Key Takeaways and What to Do Next
FROM THE LIBRARY ›
References
Editorial StandardsCorrectionsMedical DisclaimerAbout Our ContentAffiliate DisclosureSite Map