Strategy is a decision-making framework that defines what to do and what to skip; the article gives a crisp definition and three actionable models-Rumelt's Kernel, Playing to Win, and Three Horizons-to turn vague "be more strategic" feedback into concrete trade-offs.
Strategy isn't a roadmap; it's a framework that narrows the decision aperture by spelling out what you'll do and what you won't. The author defines strategy as an articulation of a path forward and explicit trade-offs that align actions with core objectives, and shows why vague advice like "be more strategic" fails without a concrete model.
The piece then walks through three practical frameworks. Rumelt's Kernel breaks strategy into diagnosis, guiding policy, and coherent actions, illustrated with a testing-coverage example. Playing to Win forces engineers to answer five questions about aspiration, arena, winning approach, needed capabilities, and management systems, tying technical work to business goals. The Three Horizons model maps short-term fixes, emerging opportunities, and long-term bets, showing how today's work can seed future architectural shifts.
By applying these models, leaders can diagnose bottlenecks, prioritize work that matters, and balance immediate reliability needs with longer-term scalability. The article gives concrete steps-like building end-to-end test suites or defining service boundaries-that turn strategic intent into actionable engineering plans.
Check out the full stdlib collection for more frameworks, templates, and guides to accelerate your technical leadership journey.