I read many books on leadership and strategy. Some sound lovely, but when I try to transfer them into practice, little remains beyond general appeals and a few examples.
For me, Good Strategy / Bad Strategy by Richard Rumelt is one of the most well-founded works on modern strategy. Not only because it makes clear that fine-sounding words trigger nothing. But because it describes comprehensibly the uncertainty under which strategies are created today, and because it offers something rare in this genre: a precise testing tool.
The kernel
Rumelt’s central thesis is disarmingly simple, and it reverses the usual order. A good strategy does not begin with lofty goals but with the central problem. It starts from reality — from the actual bottlenecks, from the organization’s origins and its existing potential, and derives from this which goals are achievable.
Bad strategy does it the other way around. Detached from reality, origins and resources, it formulates a pipe dream and sets goals that are not anchored in the organization’s capabilities.
What follows is not a strategy but hope with subheadings.
The kernel consists of three elements that build on one another. First, the diagnosis: an honest analysis of what is actually going on — reducing an overwhelmingly complex situation to the decisive challenge. Second, the guiding policy: the chosen basic approach. It works like a guardrail, because it explicitly rules out a great deal. Third, coherent actions: coordinated steps backed by resources.
The most famous example is Apple in 1997. Steve Jobs took over a company with dozens of product variants on the brink of bankruptcy. Diagnosis: too broad for its own strength. Guiding policy: radical focus; the portfolio shrank to four products. The actions followed coherently, from development to sales.
Just as instructive is what Jobs did not do. No growth vision, no saving the world, but surviving and, as he later told Rumelt, waiting for the next big thing.
Six questions that make the difference
1 · What is really going on? Not: what do we wish for? But: which critical challenge stands in our way? A diagnosis that hurts no one usually is not one.
2 · What is the crux? Of a hundred problems, two are decisive. Whoever tackles all of them at once tackles none.
3 · Which path do we choose, and what do we explicitly not do? A guiding policy that rules nothing out is not one.
4 · Do our actions reinforce one another? Rumelt’s model case is IKEA: catalog, self-pickup, flat packs, in-house designs. Every element supports the others. That is exactly why no one could copy it for decades.
5 · What is our next achievable proximate objective? Close enough to reach with confidence. Important enough to make everything that follows easier.
6 · Where does concentrated effort create the breakthrough? The question is not where we could get better everywhere.
How to recognize bad strategy
Fluff. Jargon that disguises banality as depth. Rumelt’s favorite example: a major bank announced “customer-centric intermediation” as its fundamental strategy. Translated: we are a bank.
Failure to name the problem. Rumelt dissects the case of International Harvester — a strategy paper full of financial targets in which the poisoned relationship with the workforce, the real problem, did not appear. The company no longer exists.
Goals mistaken for strategy. A CEO whose plan consisted of the 20/20 formula — twenty percent growth, twenty percent margin — plus an appeal to try harder. Asked how that was to be achieved: silence.
Wish lists. Rumelt cites a municipal plan with forty-seven strategies and one hundred seventy-eight actions.
From the book to the retreat
One objection I hear again and again in practice: “We need more facts before we can plan implementation.” This is exactly where many fail. There is never complete certainty in strategic questions; whoever waits for it does not decide but postpones.
When I accompany leadership teams, I therefore look for something else: that there are no agonizing, separate discussions about vision, mission and values whose results hang in the room like loose ends, but that the strategy can be told as a coherent narrative. It connects three things: the target picture, the key lines of implementation and the next milestones. Only then do leaders and employees understand the context, the why and the priorities.
Rumelt’s triad translates directly into workshop practice.
For the diagnosis, the pre-mortem has proven its worth: you travel in your mind to a future in which the strategy has failed and ask why. That puts the real bottlenecks on the table, and demands management’s courage to acknowledge them openly.
The guiding policy becomes the target picture with clear acceptance criteria: how do we recognize in reality that we are succeeding? This does not mean lagging metrics but principles that set focus, and leading indicators that show in time where one stands in the competition.
The coherent actions are best conveyed as a roadmap: the next quarters worked out in concrete interim goals, everything beyond as building blocks along the way.
Two questions for your leadership team
Do your core strategic statements pass the negation test?
Can your strategy be told as a story in five minutes — with a target picture, paths and next milestones?
Where this approach reaches its limits
Rumelt’s triad is a testing tool, not a generating tool. It reliably tells you that a strategy is bad. It does not tell you which one would be good.
And the negation test has an uncomfortable side effect: it reliably devalues part of what is in your current strategy paper. Anyone who applies it in the leadership team should clarify beforehand whether the group is ready to face the result. Otherwise the test wins and the strategy stays as it was.