Business Report Rules

A business report exists to drive a decision or an action, not to document everything the analysis touched — every choice below follows from optimizing for a reader who may only read the first paragraph.

Conclusion-first (BLUF)

• Put the bottom line up front: the main finding and its implication in the first paragraph, before any methodology or background. A reader who stops after paragraph one should still know what happened and what it means.
• State the conclusion as a specific claim ("Q3 churn rose 4 points, driven almost entirely by the SMB segment"), not a topic label ("Q3 churn analysis"). A topic label makes the reader do the work the report was supposed to do.
• Resist the chronological instinct to walk the reader through the analysis in the order you did it. The order you discovered things is rarely the order that serves the reader — lead with what you found, then show how you know.

Data with context

• Never present a number without a comparison point — versus last period, versus target, versus a benchmark. "Revenue was $2.1M" is not yet informative; "$2.1M, up 12% QoQ and 8 points ahead of forecast" is.
• Flag data limitations and caveats next to the number they affect, not buried in a methodology appendix — a reader making a decision needs to know in the moment whether a figure is provisional, estimated, or based on a small sample.
• Round to the precision that matters for the decision. Three decimal places on a directional metric signals false precision and actually undermines credibility rather than building it.

One chart, one message

• Every chart should have a single, statable takeaway that could be its caption — if you can't summarize what a chart proves in one sentence, it's probably trying to show too much and needs to be split or simplified.
• Title each chart with the finding, not the variable ("SMB churn doubled after the pricing change," not "Churn by segment") — the title should let a skimming reader absorb the point without stopping to interpret the chart itself.
• Cut chart types to what the data actually needs. A bar chart for a comparison, a line for a trend, a table for exact values a reader will look up — resist stacked or dual-axis charts that require a legend and a few seconds of decoding to read correctly.

Recommendation sections

• Separate "what we found" from "what we recommend" as distinct sections — findings are what the data shows; recommendations are a judgment call built on top of them, and conflating the two hides where the judgment enters.
• State each recommendation as a specific action with an owner and, where relevant, a timeframe — "We should improve onboarding" is not a recommendation; "Redesign the day-1 email by end of Q4, owned by Growth" is.
• Where a recommendation has real trade-offs or a credible alternative, name the alternative and say why the recommended path wins — a recommendation presented with no visible alternative reads as less credible, not more decisive.

Appendix discipline

• Move anything a typical reader won't need — full data tables, detailed methodology, secondary cuts of the analysis, raw survey verbatims — to an appendix, referenced by pointer from the main body ("see Appendix B for the segment-level breakdown").
• Keep the main body readable end-to-end without appendix lookups required — the appendix is for the reader who wants to verify or dig deeper, not for content the main argument actually depends on.
• Label appendix sections clearly and in the order they're referenced, so a reader following a pointer from the body finds it immediately rather than scanning a dozen untitled tables.

Structure and length

• Default structure: BLUF summary, key findings (2-4, each with supporting data), recommendations, appendix. Don't add sections that don't map to a reader need just because a template has a slot for them.
• Match length to stakes and audience seniority — a weekly ops update might be half a page; a quarterly strategy review might run several pages — but in both cases the BLUF and recommendations should be readable without scrolling past the fold.
• Write the summary last, after the findings and recommendations are drafted, even though it appears first — an accurate summary requires knowing the whole shape of the report, which you don't have until you've written the rest.

What to avoid

• Don't include a chart or data cut "for completeness" that doesn't serve the report's conclusion — every element should earn its place by advancing the reader toward the recommendation, not just because the analysis happened to produce it.
• Don't hedge every finding into vagueness to avoid being wrong — a report that says "results may suggest a possible trend" everywhere is not more careful, it's less useful; state findings plainly and flag genuine uncertainty specifically where it exists.
• Don't bury the recommendation inside the findings narrative — a reader should be able to jump straight to a labeled recommendations section without reconstructing it from the analysis prose.