The Kernel of Good Strategy
Every real strategy has three parts: a diagnosis, a guiding policy, and coherent action.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 3
- Confidence
- 95%
Rumelt argues that strategy is not a set of analytical tools or high-level aspirations but a specific structure he calls 'the kernel' with three inseparable elements. A diagnosis names what is really going on, a guiding policy chooses how to deal with it, and coherent actions actually carry it out. If any one of the three is missing, what you have is not a strategy but something else (a goal, an ambition, or fluff).
Origin
Developed by Richard Rumelt and introduced in his 2011 book 'Good Strategy Bad Strategy.' He describes arriving at it around 2004 after decades of teaching analytical frameworks (BCG matrix, five forces, learning curve) and concluding they were tools for analysis, not strategy itself.
Core principles
- 01Strategy is always about dealing with a challenge, issue, or problem, not about mission or vision
- 02A diagnosis is a decision about what to pay attention to, since no one can understand all of reality
- 03A guiding policy chooses what you will and won't do; it is not a grab-bag of 17 priorities
- 04Coherent actions must not fight one another (e.g. burn less oil while importing more oil is self-contradictory)
- 05All three elements must be present or it is not a strategy
How to run it
- 1
Diagnose the situation
Form a rich understanding of what is actually going on. Decide what to pay attention to and generate hypotheses about how the pieces connect. This is deliberately selective because you cannot comprehend all of reality.
Pro tip Go beyond simple 2x2 diagrams; the world is more complicated than two-by-twos.
Watch out A statement like 'we are not growing fast enough' is a value or ambition, not a diagnosis. A diagnosis explains why.
- 2
Set a guiding policy
Decide how you will deal with the diagnosis: what you are doing and, just as importantly, what you are not doing. This is the core of strategy and must narrow the field.
Pro tip 'Priority' means what comes first, not a grab-bag of everything you wish would happen.
Watch out A client with 17 priorities has no policy; that is a wish-list, the opposite of a guiding policy.
- 3
Commit to coherent action
Specify the actual steps you will take. They must address the diagnosis and reinforce rather than contradict each other.
Pro tip Keep the actions few; concentration is the fundamental source of power in strategy.
Watch out Do not pursue self-contradictory goals like simultaneously maximizing growth and return on equity when they trade off against each other.
In the wild
Rumelt uses Microsoft as a concrete illustration: it diagnoses the challenge of how to adapt to AI, forms a guiding policy of investing in a major AI leader and incorporating it into its search engine, then takes coherent actions to actually build those capabilities.
→ A clean example of the three elements aligned; Rumelt notes 'it's not rocket science' yet most organizations fail to do it.
Rumelt describes companies whose 'future' is stated as investing in Industry 4.0, robotics, AI and computer vision, but whose actions amount to buying three companies with no synthesis.
→ He calls this 'strategic ass without any synthesis' — actions with no coherent guiding policy behind them.
Common mistakes
Treating analytical tools as strategy
Five forces, matrices, and learning curves help you comprehend a situation but are not themselves a strategy.
Skipping the diagnosis
Jumping straight to actions without deciding what the real problem is leaves the strategy hollow and easily hijacked by politics.
Is it for you?
Best for
Leaders and product managers who need to convert a vague mandate into a real, defensible strategy document
Not ideal for
Situations that only require a routine operating plan or checklist where the challenge is already well understood and uncontested
From the transcript
“strategy is always about dealing with an issue a challenge a problem”
“it's a form of dealing with challenges”
“a rich diagnosis of the situation but then a guiding policy”
“the most important part of a strategy the part that it's so easy to leave out because people like to think of strategy as this…”
“if anyone in three is missing something it's not really a strategy it's something else”
From the episode
Good Strategy, Bad Strategy
Richard Rumelt