Book Review: Good Strategy / Bad Strategy (by Richard Rumelt)

Key Takeaway: Richard Rumelt's Good Strategy / Bad Strategy argues that most of what passes for strategy is not strategy at all, but a mix of ambition, vision, financial targets, and activity lists that never add up to a coherent way of winning. He offers a 3-part kernel for real strategy: a diagnosis of the situation, a guiding policy for handling it, and coherent actions that follow. His most lasting contribution is naming the 4 hallmarks of bad strategy: fluff, failure to face the challenge, mistaking goals for strategy, and bad strategic objectives.

If you've sat in many strategy meetings, you've probably had this feeling: the deck is impressive, the language is confident, the leadership team is aligned, and something is wrong. The discussion is moving forward but it isn't actually addressing anything. The vision is inspiring, the targets are bold, the action items are numbered. And nobody can quite tell whether the company has a strategy or just an unusually well-formatted aspiration.

That feeling has a name. Richard Rumelt named it. The book is called Good Strategy / Bad Strategy, published in 2011, and it is the single best diagnostic tool I know for the gap between what most companies call strategy and what strategy actually is.

This is the first post in a new Perdoo series of book reviews. We're starting with this one because, in a real sense, all of the other work we do on strategy execution, OKRs, and goal-setting only matters if the underlying strategy is real. If your strategy is fluff, no execution system in the world will save it. Rumelt is the writer who made that case clearer than anyone else.

Who is Richard Rumelt, and why does this book carry weight?

Strategy as an academic field has plenty of writers. Rumelt is, by reasonable consensus, the most important one alive. McKinsey Quarterly has called him "strategy's strategist" and "a giant in the field of strategy." The Economist has profiled him as one of the 25 most influential living thinkers on management. The book itself was shortlisted for the Financial Times and McKinsey Business Book of the Year Award in 2011.

His credentials are unusual for the genre. Trained as an electrical engineer at UC Berkeley (BS, MS), he then earned a doctorate at Harvard Business School in 1972 with what may be one of the most important strategy dissertations ever written: he was the first person to establish a statistical link between corporate strategy and profitability, finding that moderately diversified firms outperform highly diversified ones. That finding has held up across more than three decades of subsequent research. He spent most of his career at UCLA Anderson, served as president of the Strategic Management Society, and has consulted for Shell, Apple, Microsoft, Intel, AT&T, the US Department of Defense, and many others.

In other words, this is not a TED-talk strategist. Rumelt has done both the academic work and the consulting work, and Good Strategy / Bad Strategy is the distillation of about 40 years of both. That matters because the book reads with a level of pattern recognition that someone who's only consulted, or only researched, simply couldn't produce.

The big idea: most "strategy" isn't strategy

The book's central claim is uncomfortable. Rumelt argues, with a remarkable amount of evidence, that the term "strategy" has been so thoroughly hollowed out by corporate use that most documents and meetings labeled "strategic" are not strategy at all.

In his own words, from the companion essay he wrote for McKinsey Quarterly:

"Despite the roar of voices equating strategy with ambition, leadership, vision, or planning, strategy is none of these. Rather, it is coherent action backed by an argument."

This is the whole book in two sentences. Strategy isn't a feeling. It isn't a vision. It isn't a goal. It isn't a plan to do many things. Strategy is a coordinated way of dealing with a specific obstacle, supported by a diagnosis that explains why this approach should work.

Once you've internalized that definition, you start seeing bad strategy everywhere, and Rumelt's book is essentially a guided tour of how to spot it.

The 4 hallmarks of bad strategy

The book's most-cited contribution is its identification of the 4 telltale signs that what's being called strategy isn't actually strategy. These are worth knowing by heart.

1. Fluff.

Rumelt's definition: "a restatement of the obvious, combined with a generous sprinkling of buzzwords that masquerade as expertise." His favorite example in the book is a major retail bank whose strategy was described as "customer-centric intermediation." Rumelt's translation: intermediation means accepting deposits and lending money out. In other words, the bank's strategy is to be a bank. The buzzword "customer-centric" couldn't be backed up with any actual policies that made the bank more customer-centric than its competitors. Once you've seen this pattern named, you can't unsee it. Every strategy deck you read for the rest of your career will have a "customer-centric intermediation"-moment.

2. Failure to face the challenge.

A strategy is a way of dealing with an obstacle. If the obstacle is never explicitly named, the strategy can't actually be evaluated, because there's no benchmark for whether the proposed approach is the right one. Rumelt's most painful example here is International Harvester, which in 1979 produced a 5-part strategic plan spanning hundreds of pages, projecting market share growth from 16% to 20% and significant profit improvements. The plan never mentioned the actual challenge facing the company, which was that Harvester had the worst labor relations in US industry and its production facilities were grossly inefficient. The plan was extensive and detailed. It addressed none of the things that mattered. Within a few years, Harvester collapsed and was broken up. The agricultural-equipment business was sold to Tenneco. The truck business was renamed Navistar.

3. Mistaking goals for strategy.

This one is the most common, and probably the most important for OKR practitioners to understand. Rumelt is blunt: "Many bad strategies are just statements of desire rather than plans for overcoming obstacles." A target like "grow revenue 20% per year" or "achieve 20% operating margin" is not a strategy. It's a goal. The strategy is the answer to how you're going to do that, given the specific obstacles in your way and the specific advantages you can exploit. Rumelt is the writer who first crystallized why the confusion between the two is so destructive. He devotes an entire chapter to what he calls the "20/20 plan," a real client interaction where a CEO insisted that committing harder to a 20% growth and 20% margin target was itself the strategy. Rumelt's response, distilled: a stretch goal is a wish. A strategy is an argument about why a specific approach will move you toward that wish, given the conditions you actually face. Without the argument, you only have a wish.

4. Bad strategic objectives.

Rumelt distinguishes between long-term ambitions (which are fine and even necessary) and immediate strategic objectives (which need to be tightly scoped, achievable with current resources, and connected to the diagnosis). When strategic objectives are vague ("become a leader in our market"), impossibly broad ("transform our culture"), or unconnected to the actual challenge, they're bad strategic objectives. His example here includes a mayor's small-city strategic plan that contained 47 strategies and 178 action items. Action item number 122, he notes drily, was "create a strategic plan." The mayor had been given a list so long and so unfocused that the planning process had quietly recommended itself as one of its own deliverables.

These 4 patterns are not separate failures. They're often combined. The plan you're staring at has fluff, doesn't name the actual challenge, mistakes its growth target for a strategy, and has 47 strategic objectives that don't connect to anything. Rumelt's gift is letting you finally see the pattern.

The kernel of good strategy

If those are the 4 hallmarks of bad strategy, what does good strategy look like? Rumelt offers a 3-part structure he calls the "kernel" of good strategy. This framework is the book's most positive contribution.

1. A diagnosis.

"An explanation of the nature of the challenge," in Rumelt's words. A good diagnosis simplifies the overwhelming complexity of the situation by identifying which aspects are actually critical. This is where most strategy work fails. Leaders skip the diagnosis because it's the hard, ambiguous part. They jump straight to the policy and the action items. But without a diagnosis, there's nothing to evaluate the policy against. Dave Kellogg, in his thoughtful review of the book, goes so far as to say "80% of strategy is the diagnosis." I think that's about right.

2. A guiding policy.

An overall approach for dealing with the obstacles the diagnosis identified. Rumelt uses an analogy that's become well-known: a guiding policy is like the guardrails on a highway. It directs and constrains action without fully specifying every move. The policy says what kind of approach you're taking. The actions will say what specifically you're going to do.

3. Coherent actions.

A set of coordinated steps that implement the guiding policy. The word "coherent" is doing a lot of work here. Rumelt is emphatic that the actions have to fit together, reinforce each other, and follow from the policy. A list of activities that don't cohere isn't a strategy. It's a to-do list.

Nvidia

His most-cited illustration of the kernel is Nvidia's 1995 turnaround:

• The diagnosis: "We are losing the performance race."
• The guiding policy: "Release a faster, better chip three times faster than the industry norm."
• The coherent actions: Form three development teams on overlapping schedules, invest in simulation and emulation to avoid fabrication delays, and reclaim driver development from the branded board makers.

Each action ties to the policy. The policy ties to the diagnosis. Over the following decade, Nvidia went from struggling start-up to industry dominant, while its main competitors (3Dfx, Intel's discrete graphics business) exited or went bankrupt.

The chapter that haunts every executive: Digital Equipment Corporation

There's one chapter in the book that I think every executive should read regardless of whether they read the rest. It's Rumelt's account of a 1992 strategy session at Digital Equipment Corporation (DEC), where he was sitting in.

DEC had been losing ground for years to 32-bit personal computers. Three senior executives proposed three different strategies. "Alec" argued DEC should keep integrating hardware and software into computer systems (the "Boxes" strategy). "Beverly" argued the company's only distinctive resource was its customer relationships, and it should pivot to a "Solutions" strategy that solved customer problems. "Craig" argued the heart of the industry was semiconductors, and DEC should focus on designing better chips.

These were not compatible strategies. Each represented a different bet about what DEC actually was and where it should compete. The CEO, Ken Olsen, did what most CEOs do when faced with this kind of disagreement among smart, senior people: he asked the group to reach consensus.

What emerged from that consensus exercise, Rumelt records, was this:

"DEC is committed to providing high-quality products and services and being a leader in data processing."

He calls this "fluffy, amorphous statement" exactly what it was: not a strategy, but a political outcome reached by people who couldn't agree on which interests to forego. It was the safest sentence that all three executives could sign their names to. It said nothing.

Ken Olsen was replaced in June 1992. The new CEO, Robert Palmer, eventually picked one strategy (Chips). One point of view had finally won. But by then, Rumelt writes, "it was five years too late." DEC was acquired by Compaq in 1998. Compaq was acquired by Hewlett-Packard three years later. A computing pioneer was gone.

The DEC chapter is uncomfortable reading because it implicates so much of how senior leadership teams actually operate. The instinct to "align," to find the sentence everyone can agree to, to avoid the political cost of choosing one path over another, is the same instinct that produces bad strategy across thousands of companies every year.

Rumelt is direct: strategy involves focus, and focus means choosing some directions and explicitly setting others aside. When leadership teams refuse to choose, they get the DEC outcome.

What the critics get right

Even great books have honest critics, and Good Strategy / Bad Strategy is no exception. Dave Kellogg, a longtime SaaS CEO and one of the more thoughtful executive bloggers writing on strategy, calls it "by far my favorite book on strategy" and calls Rumelt's critique of bad strategy "borderline therapeutic." But he also points out a real limitation: the book is significantly stronger at tearing apart bad strategy than at helping you build good strategy yourself. Kellogg even suggests, only half-jokingly, that the book "might have been better titled Bad Strategy, Good Strategy."

He's right. The diagnostic apparatus is extraordinary. The constructive apparatus is thinner. Rumelt's kernel framework is conceptually clear but, as Kellogg notes, "does not even attempt to outline a process whereby an executive team can work together to devise a good strategy." You finish the book with a much sharper eye for when something is fluff or a wishful goal masquerading as a strategy, but with less help on how to actually facilitate the strategy conversation in your own company.

A second fair criticism: the framework can be genuinely hard to apply, because if the diagnosis is wrong, "the whole strategy collapses along with it," as Kellogg puts it. A weak diagnosis is worse than no diagnosis, because it gives the strategy false confidence. Rumelt acknowledges this risk, but doesn't fully address how a leadership team can pressure-test its diagnosis before committing to it.

A third caveat, which Rumelt himself has implicitly conceded by writing a follow-up book in 2022 (The Crux): the original book leans heavily on competitive threat as the framing for what makes a strategy necessary. That's a Silicon Valley-friendly framing in some ways and an awkward fit in others. Some industries don't have a single overwhelming competitive threat at any given moment. They have a slowly building obstacle, a regulatory shift, a customer-base shift, or an internal capability gap. Rumelt's framework handles these, but the book's most vivid examples lean toward dramatic competitive showdowns.

None of these criticisms invalidates the book. They sharpen its use. You read Good Strategy / Bad Strategy primarily as a diagnostic, then you supplement with other tools (Rumelt's own The Crux, or any decent strategy facilitation framework) for the constructive work.

Why this is a must-read for executives

You will start spotting bad strategy in your own company. This is the immediate payoff, and it's substantial. Within two weeks of finishing the book, you'll be sitting in a strategy meeting and you'll suddenly notice that the discussion has spent 90 minutes on goals, ambition, and language without ever naming what the actual obstacle is. You'll recognize fluff in your own decks. You'll recognize fluff in your competitors' decks. This pattern recognition is the highest-leverage thing the book gives you, and most executives don't have it because they've never seen the patterns named.

You'll stop confusing OKRs with strategy. This is the version of the lesson that matters most for Perdoo customers and anyone running a serious goal-setting program. OKRs are an execution layer, not a strategy layer. Rumelt's distinction between strategy and goals is one of the cleanest in the literature, and it underpins why we keep insisting that you set strategy first and then operationalize it with OKRs. A company with a clear strategy and decent OKRs will outperform a company with great OKRs and no strategy, every time. The OKRs in the second company will be exquisitely measured, beautifully cascaded, and they'll move the business nowhere because there's no strategic argument underneath them.

You'll get a clearer sense of when not to act. Most strategy books push leaders toward action. Rumelt's book is unusual in how often it counsels patience, restraint, and refusal. Strategy involves choosing what not to do, what not to pursue, what not to optimize. The DEC chapter is the most vivid argument for this discipline, but the theme runs throughout. If your culture has a bias toward action, this book is a useful corrective.

You'll understand why most strategic plans fail. It's not always that the strategy was wrong. Often, there was no strategy, only a budget and a list of activities and a value statement, sometimes poorly translated into the execution layer of the organization. Other times, the strategy was sound but only the top of the company was ever executing it, so the strategy stayed in the boardroom and never reached the front line. Rumelt's framework gives you the language to diagnose the first failure mode. The second failure mode is what good execution systems are built to prevent.

The writing is unusually good for a strategy book. Most strategy books are turgid. Rumelt's prose is clean, conversational, and dryly funny in places. The Chad Logan chapter (the 20/20 CEO) is genuinely entertaining to read, even as it eviscerates a common pattern of executive thinking. The book is around 320 pages and reads quickly.

The verdict

Good Strategy / Bad Strategy is the strategy book most executives have heard of, often own a copy of, and rarely have actually read. The gap between thinking you've internalized it and actually having internalized it is enormous. You should close it.

If you can only take one thing from it, take this: a strategy is a coherent set of actions backed by an argument about how those actions will deal with a specific obstacle. If you can't articulate the obstacle, the argument, and the actions in plain language, you don't have a strategy. You have one of the 4 substitutes Rumelt names, all of which look like strategy and none of which are.

If your company is running on goals without strategy, on vision without diagnosis, or on a 47-item action list without an underlying argument, this book will name the disease. The naming is the first step toward fixing it. And the fix, properly done, is what unlocks every other system in your organization, including OKRs, KPIs, planning rhythms, and quarterly reviews.

We give it the strongest possible recommendation. It's the book to start with if you're going to read one strategy book in your career.

[fs-toc-omit]How Perdoo helps you connect strategy to execution

Rumelt's book is the diagnostic. The next question is what happens once your strategy is clear. That's where the day-to-day execution layer needs to do its job, by turning the guiding policy and coherent actions into the specific Objectives, Key Results, and Initiatives that the rest of the company will actually work on. Perdoo is built for exactly this translation. Strategic Pillars at the top of your Strategy Map, OKRs for the time-bound changes you're driving, KPIs for the maintenance metrics that show whether the strategy is working, and Initiatives for the actual projects underneath. Each one tied to a parent goal, visible to the whole company by default. If your strategy is real, start for free or request a demo and give it a system that will translate it into action across every team.