Manu's Vault

Good-Strategy-Structure

4 min readStatus:Seedling

Most engineering “strategies” are not strategies at all. They are lists of aspirations (“we will be world-class”), goal inventories (“achieve 99.9% uptime, reduce latency by 30%, ship two major features per quarter”), or vision statements dressed up as plans. They tell you where you want to go but say nothing about why getting there is hard or how constraints shape the path. Richard Rumelt calls this bad strategy, and Will Larson applies the diagnosis directly to engineering leadership.

Rumelt’s Kernel of Good Strategy

A good strategy has three interlocking parts — what Rumelt calls the kernel:

  • Diagnosis: A clear-eyed assessment of the specific challenge. Not “we need to scale” but “our deployment pipeline averages 4 hours because 47 integration tests flake on every run, blocking daily releases.” A good diagnosis names the constraint, not just the symptom.

  • Guiding Policies: Rules and principles that address the diagnosis by shaping decision-making. They define the decision space without specifying exact actions. Example: “We will not add new integration tests until flake rate is below 5%. We will invest in parallel pipeline stages as the primary path to speed.” Guiding policies say what you will and won’t do, and why.

  • Coherent Actions: Specific, sequenced steps that implement the guiding policies. Example: “Q1 — one engineer dedicated to flaky test elimination. Q2 — implement parallel build stages. Q3 — measure and iterate.” Actions must be mutually reinforcing, not pulling in different directions.

What Bad Strategy Looks Like in Engineering

  • Aspirations masquerading as strategy: “We will be the best engineering team in our space” — no diagnosis, no policies
  • Goal lists with no connective tissue: each objective unrelated to a structural constraint
  • Guiding policies that don’t address the diagnosis: the policy solves a different problem than the one named
  • Incoherent actions: teams pulling in opposite directions simultaneously
  • Conflating strategy with vision — vision describes the destination; strategy describes how to navigate the specific terrain between here and there

The Most Commonly Missed Step: Diagnosis

Larson observes that most engineering strategies fail at the Diagnosis step, not the action-planning step. Engineering managers are optimists by training — they diagnose “we need more resources” rather than naming the structural obstacle. A good engineering strategy must name the specific constraint: the architectural coupling, the process bottleneck, the knowledge concentration, the organisational misalignment.

Without an honest diagnosis, guiding policies float free of reality, and coherent actions become impossible to sequence.

Applying the Kernel

To construct a good engineering strategy:

  1. Name the single hardest thing — the constraint that makes the goal difficult
  2. Define 2-4 policies that directly address that constraint
  3. Sequence concrete actions that are consistent with each other and with the policies
  4. Verify: do the actions implement the policies? Do the policies address the diagnosis?

Sources

  • Rumelt, Richard P. (2011). Good Strategy Bad Strategy: The Difference and Why It Matters. Crown Business. ISBN: 978-0-307-88623-1.

    • Chapters 1-5 define the kernel of strategy and identify the hallmarks of bad strategy; primary theoretical source for this note

  • Larson, Will (2019). An Elegant Puzzle: Systems of Engineering Management. Stripe Press. ISBN: 978-1-7322651-8-9.

    • Chapter 5.4 applies Rumelt’s kernel to engineering leadership; adds the observation that Diagnosis is the most commonly missed step

  • Porter, Michael E. (1996). “What Is Strategy?” Harvard Business Review, November–December 1996, Vol. 74, No. 6, pp. 61-78.

    • Foundational distinction between operational effectiveness and strategy; strategy requires trade-offs and fit among activities — adjacent framing to Rumelt’s kernel

  • Mintzberg, Henry, Bruce Ahlstrand, and Joseph Lampel (1998). Strategy Safari: A Guided Tour Through the Wilds of Strategic Management. Free Press. ISBN: 978-0-684-84743-7.

    • Surveys ten schools of strategy thought; contextualises the design school (deliberate strategy formation) from which Rumelt’s kernel derives

  • Lafley, A.G. and Roger L. Martin (2013). Playing to Win: How Strategy Really Works. Harvard Business Review Press. ISBN: 978-1-4221-8739-6.

    • Presents the “five choices” strategy cascade (winning aspiration, where to play, how to win, capabilities, management systems); complements Rumelt’s kernel with a practitioner implementation model

Note

This content was drafted with assistance from AI tools for research, organization, and initial content generation. All final content has been reviewed, fact-checked, and edited by the author to ensure accuracy and alignment with the author’s intentions and perspective.

Graph View

Backlinks