The Best Book on Strategy I’ve Ever Read

Good Strategy, Bad Strategy
Richard Rumelt

Favorite Line: A great deal of strategy work is trying to figure out what is going on. Not just deciding what to do, but the more fundamental problem of comprehending the situation.

Funniest Line: To detect a bad strategy, look for one or more of its four major hallmarks: Fluff. Fluff is a form of gibberish masquerading as strategic concepts or arguments.

Overall Impressions
My god, I love this book so much. When I first started exploring strategy books, I had already written several city plans, participated in strategic retreats, and scratched out “business plans” for my department. I am a planner by trade, after all. But when the first few works failed to entice any action, I knew something was wrong. So I started reading. My first few books were unsatisfying because they were an introduction and defense of a particular kind of framework. The Balanced Scorecard or Five Forces or, within city planning, the Rational Planning Model. Only until I came across Henry Mintzberg did I start to find better books that focused on the act of planning itself rather than the means by which the act could be performed. It was the difference in reading about the heuristics of the craft versus the technique of the tools.

With that in mind, my journey culminated in the discovery of the best book ever written on the subject.

I don’t think I would have understood the importance of this book had I read it first before all those others that came before it. By learning about the different frameworks for strategy, I had the chance to question why any of those frameworks mattered at all. And to ask why one might be better than another. What does it matter if someone uses McKinsey’s Startegic Horizons or the Ansoff Matrix? Aren’t there countless examples of companies that have been successful without using any of these tools?

My skepticism started to grow. So much of this stuff felt like the diet fads that plague the nutrition industry. Rumelt’s book helped me to understand why. As he puts it,

A good strategy has an essential logical structure that I call the kernel. The kernel of a strategy contains three elements: a diagnosis, a guiding policy, and coherent action.

I wouldn’t usually fixate on this sentence while reading a book. It is the introduction of the concept, the thesis, but I would be looking elsewhere for second-layer stuff that gets at the real nuance. Not this time, however. Because strategy is one of those things that quickly becomes complicated when it desperately needs to remain simple. Our natural inclination, thanks to the overwrought frameworks and plans and efforts we all develop and call “strategy”, is to make this work very complex very fast.

Or worse yet, to ignore plain truths. So when formulating strategy and working with others to identify what it is, I’ve always struggled to find a simple, core idea to keep people grounded. Here it is. The kernel. The word is the perfect label; it’s a word we don’t use enough today outside the context of operating systems. We refer to “first principles” as a means of thinking clearly, which I fully endorse, but before first principles there is this.

But again, what’s the big deal? This feels so obvious to some, I’m sure. So what matters here is the fact that most strategic plans are not built on, or at least fail to communicate, a proper kernel. I honestly can’t think of one such document that explicitly states it. Certainly not in local government but I don’t think I’ve formally seen it elsewhere either. How wonderful would it be for Amazon or Google or GM to say “Here is our strategy based on this diagnosis, this guiding policy, and this coherent action”? Nothing more, nothing less.

I might be the only one who gets excited at the thought but this applies to all of life. We use this calculus to formulate our grocery lists. We have a diagnosis—the fridge is empty and the cupboard is bare—and we need food; we make a list of groceries based on a guiding policy—not too expensive and we’re trying to stick to a Paleo diet—and we then embark on coherent action—going to the store.

That’s strategy. And since this concept of the kernel is so foundational, it can scale to the size of any other operation. This idea of the kernel can unpack every planned action of the D-Day invasion. There will be a lot of complexity and interconnectedness when we analyze those plans but the core, again, is simple.

Which is to say that no one needs any strategic framework. We just need clear thinking. Additional elements like timeframes, goals and objectives, market segments, and the sort might be helpful but, again, it isn’t necessary. Or as Rumelt puts it:

I call this combination of three elements [diagnosis of a problem, guiding policy, and coherent action] the kernel to emphasize that it is the bare-bones center of a strategy—the hard nut at the core of the concept. It leaves out visions, hierarchies of goals and objectives, references to time span or scope, and ideas about adaptation and change. All of these are supporting players.

Supporting players. Mission statements, vision statements, gantt charts, scope. They’re all a bunch of goddamn appurtenances. Millions of hours of staff time (to say nothing of the financial cost) has been spent writing the perfect vision statement and color-code our critical paths. I should know. I’ve done a lot of it myself. I love it. But more times than not, it is creating the structure and formula for future action even as the kernel—that vital core—is still ignored.

So here is my kernel for future strategy sessions. A strategy for the strategizing:

Diagnosis of the problem: strategy documents fail to produce a genuine kernel and thus are empty documents built as a means of deciding what to do without ever figuring out what’s going on.

Guiding policy: no strategy document will be developed until a kernel is fully formed and articulated front and center. Furthermore, strategy documents may be considered wholly unnecessary if and when a kernel is fully developed and agreed-upon by the primary actors. The kernel will consist of the diagnosis, guiding policy, and coherent actions. The diagnosis may require data and experimentation to prove itself (or not) and the guiding policies might require data to illustrate the necessary proof of interdependencies between variables that we wish to change (Paleo is good because carbs are bad; carbs are bad because they lead to insulin spikes and fat gain). Coherent actions should be styled in a manner that is entirely within our control and possess as much of an economy of resources as possible.

Coherent action: for our next strategy session, we will begin with the diagnosis of the problem. Assuming we’ve chosen the problem to address. If we haven’t chosen the problem to address, we will work on that for as long as necessary. Once diagnosed, a problem will be addressed through coherent action built on the guiding policies and all will be documented in a simple, brief document. Timeframes, RACI matrices, project charters, goal statements, objectives, and other such things will be added only as needed.

There’s so much more that this book captures. The thoughts on bad strategy are just as valuable as the thoughts on good strategy. And for all of us who have suffered through long retreats and strategy sessions that kinda lead to nowhere, these measured rants are cathartic. Rumelt shares our frustration. Strategy can be so simple but no one seems to know that.

Mental Models

  • Real strategy requires the kernel: problem diagnosis, guiding policy, and coherent action.
  • A problem well-defined is half-solved —Oscar Wilde
  • Vague language is a sign of confusion and fear to choose.
  • Bad strategy is not the lack of good strategy. It is its own dangerous thing and crowds out good strategy.
  • Goals are often just statements of desire; they aren’t strategy.
  • Strategy is a plan to overcome an obstacle.
  • Great strategy honestly acknowledges the problems and provides coordinating and focused actions to overcome them.
  • Actions must coherent, direct responses to the problem and must be immediate.
  • Markets do not evolve from simple to complex; it’s the other way around.
  • Managing the bridge between goals and objectives is a leader’s top job.
  • Good strategy focuses on one, or a very few, pivotal objectives. A’la Michael Porter’s straddling problem.
  • Never label a condition as a problem. A condition, like underperformance, is a result or symptom. The problem is the cause, something deeper.
  • Strategy determines what purposes are worth pursing *and* achievable.
  • If you don’t have a competitive advantage, don’t compete.
  • Strategy is about figuring out what is going on; comprehending the situation is half the battle before deciding what to do.
  • Deep change in a system requires a deep change in diagnosis. Diagnosis is a leverage point. A’la Harai’s power of narrative; to change people, change the narrative. And Tim Ferriss’s “Let it serve you”, reframe the problem as a benefit. And Tony Robbin’s additional reframe: “Think of this as a gift” a setback that actually helped you in some other way.
  • You can use a shoe to hammer a nail but it will take a long time. Find the power tools.
  • Seek coordinated actions only when the gain is large. Otherwise, keep people in their specialty silos.
  • Certain aspects of future events are predetermined. Find out which ones.
  • Pivot points, by definition, emerge from the imbalance of a situation.
  • Understand the equilibrium, sense the imbalance, and act with the small adjustment that unleashes much larger pent-up forces.
  • Like people, an organization can only focus on a few critical issues at any one time.
  • Remember the power of the proximate objective—something close enough to feasibly achieve and even overwhelm.
  • Ignore ambitious goals; test their feasibility instead by judging the level of ambiguity about the obstacles to overcome.
  • Just improve your position. You don’t win chess by trying to to checkmate with every move.
  • A chain is only as strong as its weakest link. Either replace it or eliminate it.
  • There is little payoff in incrementally strengthening the link.
  • To assess potential, identify the limiting factors.
  • Never accept the first reasonable answer to a hard question or problem. Test it.
    Most advantages only extend so far. Look for an advantage that is sustainable, something others can’t duplicate.
  • Increasing value requires progress on at least one of four fronts: deepening advantages, broadening the extent of advantages, creating higher demand for advantaged products or services, or strengthening the isolating mechnisms that block easy replication or imitation.
  • Preventing duplication or imitation is the best defense and best generator of value. Continually improve methods, products, or services. A’la Bezos’s statement that old business was 70% marketing, 30% quality, and now it’s flipped.
  • Favor a business or service that can rapidly redevelopment and improve.
  • You can’t show your skill as a sailor when there is no wind.
  • Organizational inertia falls in three categories: the intertia of routine, cultural inertia, and inertia by proxy.
  • Inertia by routine is the result of bad management.
  • To kill cultural inertia, simplify routines, processes, and eliminate the hidden bargains, excess layers of administration, and halt nonessential operations. A’la Jobs’ quip that simplicity is the ultimate sophistication. And Tim Ferriss’s question: What would it look like if it were easy?
  • To change a group’s norms, change the alpha.
  • Inertia by proxy is dealt with by creating separation. Fragment units, eliminate their cross-subsidies and dependencies, destroy political coalitions, remove those who do not need to work in close coordination.
  • A loss of organizational coherence increases internal competition. Departments fight when they’re confused on a crowded dance floor.
  • A good strategy is a hypothesis of what will work.
  • To be innovative, or just plain effective, focus not on what is done but why.
  • Always start with why. Question it all.
  • Enlightened observation is built on the eternal question of why.
  • Never build a system so complex that you can’t understand or analyze its failure points and failure modes.
  • Avoid, forever, the inside view that “this time is different”.