Summary of "Good Strategy Bad Strategy"

Core Idea

Strategy is diagnosis + guiding policy + coherent action—not vague goals or aspirational statements
Bad strategy persists because leaders avoid hard trade-offs and hide behind fluff; good strategy concentrates resources on pivot points where small efforts yield disproportionate results
Strategic advantage comes from proprietary knowledge built through doing, tested like scientific experiments, then refined through real-world anomalies

Diagnosis: Rigorously identify the real problem, not symptoms (e.g., International Harvester's labor dysfunction, not just declining sales)
Guiding Policy: Commit to a specific approach that addresses the root challenge; sets constraints and direction
Coherent Actions: Implement tightly integrated, reinforcing tactics (Walmart's unified logistics, store placement, and supply chain worked together, not in isolation)

Fluff: High-sounding language with no substance
Mistaking goals for strategy: Revenue targets and mission statements aren't strategies
Refusing to choose: Trying to do everything dilutes focus and creates contradictory policies
Blue-sky objectives: New goals as vague and difficult as the original problem

Absorb ambiguity at leadership level—hand the organization a solvable, proximate objective (not an impossible leap)
Expect years to implement major shifts; routines, culture, and invested interests create resistance
Vertical integration works when it captures cross-functional learning impossible to gain otherwise

Frame strategy as testable hypothesis: Implement, observe results, adjust—don't treat it as divine truth
Generate 2-3 robust alternatives before committing; first insight feels good but may be incomplete
Spot anomalies: Compare actual results against refined expectations; gaps reveal opportunities competitors missed
Commit judgments to writing before discussion: Prevents hindsight bias and creates measurable improvement over time

Watch for closed-loop thinking: When insiders validate each other without external data, bubble thinking emerges—consult history, international comparisons, base rates
Break social herding: When everyone agrees (Fed, Treasury, investors), that's when independent analysis matters most
Challenge "this time is different": Real estate + easy credit = predictable disaster (1819, 1837, 1873, 1893, 2008); document the pattern, don't assume new conditions have changed human nature

Diagnose ruthlessly: Identify the actual bottleneck before proposing solutions; reframe symptoms as root causes
Make hard trade-offs: Say "no" to competing interests; focus resources where you have decisive advantage
Design for coherence: Ensure every policy reinforces others; loose coordination cancels gains
Test and iterate: Treat strategy as an experiment; build proprietary knowledge through doing and learning
Read primary evidence: Don't outsource judgment to authority; check historical patterns for recurring crises before they hit