Decision Memo

name: decision-memo
description: Structures a pending decision into a one-page memo with options, trade-offs, a recommendation and the strongest counter-case. Use when the user needs to frame a decision for an executive or committee, compare options or write a recommendation memo.

Turn a messy decision into the one-page memo that lets a principal decide in ten minutes. A chief of staff writes these to protect two scarce resources. The principal's attention and the organization's decision velocity. The memo succeeds when the principal can say "approved, option B" without a meeting and when six months later, anyone can reconstruct why the call was made.

Inputs

• The decision to be made and any real deadline (and what makes the deadline real).
• The options under consideration or just the raw situation; deriving the option set is part of the job.
• Constraints: budget, headcount, regulatory, key relationships, political realities worth writing down carefully.
• Who decides and who has to live with it.

Workflow

1. Sharpen the decision statement. One sentence, specific, answerable: not "what should we do about the data provider?" but "renew the current market-data contract for three years at the negotiated rate or move to the challenger and absorb a two-quarter migration?" If the user's framing is actually two decisions stacked together (whether to change AND which vendor), split them and say so. Bundled decisions are how committees deadlock.

2. Build the real option set. Usually two to four options. Test for the two commonly missing ones: do nothing / defer (always include it explicitly. It is what happens by default and pricing it keeps the memo honest) and the hybrid or staged option (pilot first, partial rollout) when one genuinely exists. Do not pad with straw options; a principal can smell a two-option memo dressed as three.

3. Assess each option on the same dimensions so they compare cleanly:

• What it is, in one line
• Expected benefit. Concrete and tied to something the principal cares about
• Cost and effort. Dollars, people, calendar time; figures from the user or marked "[estimate. Validate]"
• Key risks. The one or two real ones, not a boilerplate list
• Reversibility: one-way door or two-way door, stated explicitly. This is the load-bearing analysis: reversible decisions should be made fast and cheap, irreversible ones deserve the scrutiny. Flag any mismatch between the decision's treatment and its type. An irreversible call being rushed or a reversible one that has been stuck in analysis for a quarter.

4. Make the recommendation: which option and the two or three reasons that actually drive it (not every consideration; the driving ones). Then price the cost of delay in one line: what deciding 30 days later costs or forfeits. If the honest answer is "nothing," say so. That itself tells the principal how to prioritize it.

5. Write the case against. The strongest honest argument against the recommendation, stated as its best advocate would. A real argument with its own evidence, not a token risk restated. If writing it genuinely changes your mind, change the recommendation and say the case-against did it; that is the section doing its job.

6. Define what would change this. The one or two observable facts or events that should reopen the decision. This converts the memo from a one-time argument into a standing position with tripwires.

Memo format: one page

Decision required: one sentence; who decides; by when and why then. Context: three or four sentences on why this is on the table now. Options: the comparable assessments from step 3. Recommendation: the option, the driving reasons, the cost of delay. Case against: at full strength. What would change this: the tripwires.

Guardrails

• Do not manufacture false balance. If one option dominates, say so plainly but the case-against section still gets real content.
• Costs and figures come from the user or are marked as estimates needing validation.
• State constraints and political realities as facts without editorializing about the people involved; the memo may circulate.
• Keep it to one page. Supporting analysis can be referenced, not included. The discipline of the page is the product.