Diagnosis — naming the actual challenge
Most strategies fail at the first step. "Increase market share" isn't a diagnosis; it's a goal. The diagnosis is why market share is what it is now — the structural cause. Honest diagnoses are uncomfortable; aspirational goals are comfortable. Strategy work is choosing the uncomfortable path.
Guiding policy — the lever
Given the diagnosis, the guiding policy is the high-level approach. "We'll lean into developer-first" is a guiding policy. "We'll be the easiest tool to adopt" is a guiding policy. It rules things in and rules things out.
Coherent action — what we actually do
The actual moves: which products to build, which markets to enter, which features to ship. Coherent means the actions reinforce each other and the guiding policy. Incoherent action is "we'll be developer-first AND we'll add a no-code builder for non-developers AND we'll target enterprise IT." All three rules-in. That's not strategy; that's wish-list.
↳ the bad-strategy tell
What PMs do with strategy
PMs translate strategy into product decisions. If the strategy is "developer-first," the roadmap reflects that — features that delight devs ship; features that delight end-users-who-aren't-devs are deferred. If the roadmap contradicts the strategy, one of them is wrong.