I have written and reviewed a lot of monthly performance packs, for Country Heads, regional leaders, private equity investors and Boards. The numbers are rarely the problem. The usual problem is that a reader finishes the page without knowing what changed, why, or what is needed from them.
Start with the answer
The first line should say what happened and how much it matters: "Net revenue is $2M behind forecast, mainly on lower volume in two markets, and the full-year outlook is at risk." A leader who reads nothing else should still have the main point.
Explain drivers, and skip the arithmetic
"Gross margin fell 40 basis points" repeats the table. A driver tells the reader something new: mix moved toward lower-margin channels, or a promotion ran longer than planned. Size each driver, so the reader sees which one matters most.
Separate risks from opportunities
A forecast is a judgement. I list risks and opportunities against it, each with an amount and an owner. It shows the range of possible outcomes and opens a useful conversation about which ones the team can influence.
End with a decision or an action
Every pack should close with what happens next: a decision needed, an action with a named owner, or a clear statement that no action is required. Commentary with no next step is only a description.
Three rules for the writing
- Use the reader's words. Say "discounts" if your audience thinks in discounts.
- Keep the same order every month, so people know where to look.
- Cut anything that would not change a decision.
Generative AI can now draft much of this from a variance table, and I think that is a good use of it. The structure still has to come from finance, because knowing what matters to this reader, this month, takes judgement about the business.
Newer: Where AI earns its place in FP&AOlder: Making rolling forecasts useful