Skip to main content

Strategy Memo Lab

Produce a real 2-page engineering strategy using Rumelt's structure. The lab is the artifact, not the reading. Budget: 2-3 focused hours.

Retrieval Prompts​

  1. State Rumelt's three-part kernel from memory.
  2. What is the difference between a guiding policy and a slogan?
  3. Name the sections that make a strategy document operational (scope, anti-scope, tradeoffs, constraints, review triggers).
  4. Why is an engineering strategy not the same thing as a roadmap?
  5. What should a Later item be filtered against before it is added?

The Problem​

Choose one of the following hypothetical problems, or a real one from your work:

  • A. Your platform team ships one feature per quarter while four product teams are blocked on it. Headcount will not grow. Management has asked "what is the plan?"
  • B. Your company runs 180 services; the production incident rate is up 40% YoY; the on-call engineers are quitting. "Reliability strategy" is on the all-hands agenda next month.
  • C. A new regulation will require end-to-end audit logging across every API by next January. Today, about 25% of APIs have auditable logging. No one is funded to do this work.
  • D. A real problem from your current team, with stakes similar to A/B/C.

Compare and Distinguish​

Separate these pairs clearly before drafting:

  • Diagnosis vs symptom list. The diagnosis names the crux. Symptoms are the things you see.
  • Guiding policy vs vision statement. Policy rules something out. Vision does not.
  • Coherent action vs wish list. Coherent actions do not make sense under a different policy.
  • Scope vs anti-scope. Anti-scope names things that plausibly could be included and deliberately are not.
  • Constraint vs tradeoff. Constraints are imposed by the world. Tradeoffs are chosen by the strategy.

Common Mistake Check​

Identify what is wrong with each of these strategy excerpts:

  1. "Diagnosis: Our platform team has too much work." (symptom, not a diagnosis)
  2. "Guiding policy: We will be customer-obsessed." (slogan - rules out nothing)
  3. "Coherent actions: Invest in documentation, improve testing, increase velocity, and raise quality." (wish list with no policy backing)
  4. "Tradeoffs: This strategy improves quality, reduces risk, increases velocity, and reduces cost." (all positive - not tradeoffs)
  5. "Anti-scope: This strategy does not cover HR policy." (plausibility test fails - nobody would think it did)

The Memo​

Write a 2-page engineering strategy document with all of these sections. Target lengths in parentheses are guidance, not rules.

  1. Title and one-line summary (1 line)
  2. Diagnosis (100-150 words) - name the crux, not the symptoms. One theory of the problem.
  3. Scope (50 words) - which teams, systems, time window.
  4. Anti-scope (50-80 words) - at least three items; each must pass the plausibility test.
  5. Guiding policy (80-120 words) - the angle of attack. Must rule out at least one plausible alternative.
  6. Coherent action (200-300 words) - 3-5 actions with owners or owner-roles. Each must be derivable from the guiding policy.
  7. Tradeoffs (80-120 words) - at least three, each in the form "we give up X to get Y."
  8. Constraints (50-80 words) - budget, hiring, regulation, dependencies.
  9. Rollout (50-80 words) - how does this start and end?
  10. Review triggers (40-60 words) - at least two observable conditions.

Self-Checks​

Before calling the memo done, verify:

  • The diagnosis is one specific theory, not a list of symptoms.
  • The guiding policy rules out at least one plausible alternative.
  • Every coherent action is derivable from the policy. Try rewriting the policy as its opposite - do the actions still fit? If yes, revise.
  • At least three tradeoffs are named in "X for Y" form.
  • The anti-scope contains items that a reasonable reviewer could have expected to see in scope.
  • The review triggers are observable, not aspirational.
  • The whole thing is 2 pages or less. If over 3 pages, it is doing more than one thing.

Critique and Rewrite​

After finishing the first draft, trade memos with a peer (or give yourself 24 hours and re-read cold). For each memo:

  • Score diagnosis, guiding policy, coherent action each on 1-3 (weak / present / load-bearing).
  • Identify the single weakest section. Rewrite it.
  • Stress-test: rewrite the guiding policy as its opposite. Which actions still fit? Those are contamination. Remove or re-justify.

Evidence Check​

This lab is complete when you can:

  • produce a 2-page strategy memo in under 3 hours of focused work
  • name the diagnosis in one sentence without reading the document
  • defend the anti-scope list against "why not include X?" for at least three plausible X's
  • name one action you removed during rewrite because it did not derive from the policy