The decision placemat and the evidence behind it

Why we build the decision layer first

Our starting point for reporting redesign is the decision placemat. It brings the question, viable options, recommendation, uncertainty and accountable owner into a compact briefing. The evidence stays available behind it. We design the relationship between the two before polishing either document.

We use a simple traceability check. A reader should be able to challenge a recommendation, find the evidence supporting it and see what would change the conclusion. If the briefing is confident but the route to its evidence is unclear, the design is incomplete. If the evidence is thorough but nobody can identify the decision, it is incomplete in a different way.

The distinction reflects our operational perspective. The receiving leader has to act within a particular authority, deadline and consequence. We therefore organise the output around that situation rather than the sequence in which the analysis happened. The seven elements below are a working structure, adjusted to the decision and any formal reporting requirements.

Start with the question the recipient must answer

A long report can be necessary. Assurance work, complex analysis and regulated processes may require a detailed record. Length is not the central issue. The issue is whether the person receiving the work can identify the decision it supports and find the evidence needed to make it responsibly.

Before drafting, write the decision in one sentence. For example, the leadership team must decide whether to fund a limited pilot, change a supplier arrangement or accept a defined exposure. If the output is informational rather than decision-seeking, say that. Do not invent a decision merely to fit a template.

Then identify who has authority, what information they need, when a decision is required and what happens if it is deferred. That gives the document a purpose beyond recording the team's activity.

Separate the briefing from the evidence record

Use a short decision layer backed by an accessible evidence layer. The first explains what matters and what is being asked. The second preserves method, assumptions, detailed findings and supporting material. A reader should be able to move from a summary claim to its supporting evidence without searching through an unrelated appendix.

One page is a useful drafting constraint, not a universal rule. If the issue cannot be represented responsibly on one page, use more space. Never remove a material qualification to make a layout look cleaner.

Make comparison possible

Present alternatives on the same basis. If one option includes implementation cost and another includes only licence cost, the comparison is misleading. Use the same period, cost categories and operating assumptions. Record what you have excluded.

Distinguish a measured fact from an estimate, a judgement and a recommendation. A forecast is not a result. An interview is evidence of what someone reported, not necessarily proof of what happens across the whole organisation.

Show a range when a single number would imply false precision. Explain which assumption changes the choice. If an option works only when a particular volume, capacity or adoption level is achieved, put that dependency in the decision layer.

Use visuals to clarify relationships

A chart should answer a question. A process diagram can expose a handoff. A time series can show variation. A comparison table can reveal a trade-off. Decorative graphics that repeat the headline add little to the decision.

Label units, time periods and sources. Distinguish an actual value from a target. Do not use a green status indicator as a substitute for an explanation. A risk rating needs its assessment method and relevant uncertainty available to the reader.

Keep the supporting record suitable for the engagement's contractual, legal and professional requirements. A short briefing does not replace records that must be retained, disclosed or independently reviewed.

Test the briefing with a recipient

Ask someone who did not write the analysis to read the briefing and explain the decision, alternatives, main uncertainty and next owner. This is a comprehension check, not a performance test. Where they struggle, revise the structure rather than adding another executive summary.

After the meeting, record the decision and its conditions. Track whether actions have owners and whether new evidence triggers a review. A well-designed document is useful only if it connects to the organisation's decision process.

• Write the decision before the document.
• Compare options using consistent assumptions.
• Keep uncertainty beside the recommendation.
• Provide direct links to supporting evidence.
• Record the decision, conditions and next review.