The Decision Memo Template Your Data Team Needs (And Isn't Using)

The Output Problem

Analytics teams produce enormous volumes of work that never changes a decision. Dashboards that get glanced at. Slide decks that get skimmed. Notebooks that get bookmarked and never reopened. The problem is not the quality of the analysis. The problem is the format of the output.

Most analytical output is structured around the data, not around the decision. It presents what was found, in the order it was found, with increasing levels of detail. This is a natural way for an analyst to think. It is a terrible way for an executive to consume information. Executives do not want to explore data. They want to understand a situation, evaluate options, and decide. When the analytical output does not match that cognitive model, the insight dies in the gap between the analysis and the action.

The most effective analytical output I have ever produced was not a dashboard, a slide deck, or a notebook. It was a one-page decision memo. Four sections. No appendix. No supplementary charts. One page that contained everything a decision-maker needed to act.

The Four Sections

A decision memo has exactly four sections. Each one serves a specific function, and the order is not negotiable.

Context establishes what is happening and why it requires a decision now. This is not a history lesson or a background section that recounts every relevant metric since the company's founding. It is a tight description of the current situation and the forcing function that makes a decision necessary. Two to three paragraphs at most. If you cannot explain why this decision matters right now in three paragraphs, you have not done enough thinking to write the memo.

The context section answers two questions: What changed? Why does it matter? A strong context section creates urgency without manufacturing drama. It presents the situation with enough specificity that a reader who was previously unaware of the issue can grasp the stakes in under two minutes.

Options presents two to three courses of action, each with a description of what it involves, what it costs, and what it risks. The constraint on the number of options is deliberate. More than three options means you have not done enough pre-work to narrow the field. Presenting five options with a shrug of "here are the choices" is not analysis. It is abdication.

Each option should be genuinely viable. Do not include a clearly inferior option to make your preferred choice look better. Decision-makers see through that immediately, and it erodes trust in every memo you write after. If you have done the analysis correctly, each option will have legitimate advantages and legitimate trade-offs.

Recommendation states which option you recommend and why. This is the section where most analytical output fails, and it fails because analysts are trained to present data and leave the conclusion to the reader. A decision memo takes a position.

Taking a position does not mean being certain. It means being willing to say, based on the evidence available, this is the best course of action and here is the reasoning. If the decision-maker disagrees, the recommendation gives them something concrete to push against. That friction is productive. It forces the disagreement to be about reasoning, not about interpreting ambiguous charts.

The recommendation should be explicit about the assumptions that support it. "We recommend Option B because it has the highest expected retention impact per dollar spent, assuming the Q2 cohort behaves similarly to Q1." If the assumption is wrong, the decision-maker knows exactly which input to challenge.

Stakes describes what happens if the organization acts and what happens if it does nothing. This section forces the analyst to think about consequences, not just correlations. It also provides the most underrated input in any decision: the cost of inaction.

Most analytical presentations implicitly assume that doing nothing is free. It rarely is. A retention problem that goes unaddressed for two months has a compounding cost. A pricing issue that is deferred until next quarter has a revenue impact that is calculable. The stakes section makes these costs explicit, which often shifts the decision from "should we do this" to "can we afford not to."

Why This Format Works

The decision memo works because it matches how executives process information. Research on executive decision-making consistently shows that senior leaders prefer structured inputs that frame the decision, present constrained options, include a recommendation, and make the trade-offs explicit. Dashboards and slide decks do none of these things reliably.

The format also forces analytical rigor in a way that other outputs do not. You cannot write a credible recommendation without deeply understanding the options. You cannot write the stakes section without modeling the counterfactual. You cannot write the context section without identifying the precise forcing function that makes this decision timely. Each section creates a quality gate that catches weak analysis before it reaches the decision-maker.

There is a secondary benefit that is equally important: decision memos create a written record of the reasoning behind decisions. Six months later, when someone asks why the company chose Option B, the memo provides the full context. Dashboards do not capture reasoning. Slide decks capture a narrative that is often reconstructed after the fact. A memo captures the actual analytical logic at the moment of decision.

Common Failures

Three patterns consistently undermine decision memos.

The context section that is really a literature review. If your context section is longer than your recommendation, you are hiding behind background information instead of advancing a position. Cut it until it hurts. Then cut it again.

Options without real trade-offs. If every option has the same risk profile or the same cost structure, you have not done a differentiated analysis. The options should create genuine tension. If there were an obviously correct answer, you would not need a memo.

A recommendation without stated assumptions. Every recommendation rests on assumptions about the future. Making those assumptions explicit is not a sign of weakness. It is a sign of intellectual honesty that gives the decision-maker the information they need to evaluate your reasoning on its merits.

The Test That Matters

If your analytics team produces outputs that require a meeting to explain, the output is the problem. A well-structured decision memo should be actionable before the meeting starts. The meeting becomes a discussion of whether to accept the recommendation, not a walkthrough of what the data says.

This is the standard worth holding your team to. Not "did we produce the analysis" but "could a decision-maker act on this output without scheduling a follow-up?" The memo format makes that standard achievable. Everything else is a more elaborate way of saying "let's discuss."

Write the memo. Take the position. Make the decision possible.