The Rumelt Strategy & Constraints Model
by Will Larson • CTO at Carta at Carta
Former engineering leader at Stripe, Uber, and Calm. Author of the influential books 'An Elegant Puzzle', 'Staff Engineer', and the upcoming 'The Engineering Executive's Primer'. He writes the popular engineering leadership blog lethain.com.
🎙️ Episode Context
Will Larson discusses the shifting landscape of engineering leadership post-ZIRP, emphasizing the need to stop coddling engineers and start giving them business accountability. He breaks down frameworks for engineering strategy, systems thinking, and defining actionable company values, while sharing personal insights on writing and handling failure.
Problem It Solves
Helps leaders articulate a clear strategy when teams are misaligned or overwhelmed by too many technology choices.
Framework Overview
Based on Richard Rumelt's definition, strategy must contain a diagnosis, a guiding policy, and actions. Larson adds that effective engineering strategy often involves imposing 'boring' constraints to focus energy on what truly matters to the customer.
🧠 Framework Structure
Diagnosis: Accurately define the curr...
Guiding Policy: Determine the approac...
Action Plan: Define concrete steps to...
The Boring Principle: Constrain tools...
When to Use
When an organization lacks focus, is over-engineering solutions, or when there is debate about 'what is our strategy?'.
Common Mistakes
Confusing a goal (e.g., 'We build high quality') with a strategy, or failing to enforce the constraints defined in the policy.
Real World Example
At Uber, the strategy was 'No Cloud, Own Data Centers only,' which allowed them to launch inside China in 3 months despite geopolitical constraints. At Stripe, the 'Ruby Monolith' strategy focused devs on features rather than tooling.
The goal of good strategy is to dictate how we invest the limited capacities we have into the problems we care about.
— Will Larson