ChatGPT prompt for explaining your monthly numbers in plain language

This is the prompt we run every month to turn an accounting summary into an explanation the whole team understands. Its rule: it says what moved and by how much, but never speculates on causes you did not give it. What it cannot explain, it raises as a question rather than inventing.

The prompt to copy

Here's our monthly summary vs last month. Write a plain-language explanation for the team: what moved and by how much, top 3 things to watch, and 3 questions these numbers raise that the numbers alone can't answer. Do not speculate on causes I haven't given you. Data: {paste}

Replace the field in curly braces with your two monthly summaries, this month and last, then run.

When to use it

Every month at close, right before Playbook P10. This is the prompt that turns a line of figures into a shared read: instead of sending a table nobody reads, you send three clear paragraphs and three questions that start the right conversation.

What you get back

1. What moved month over month, and by how much, in plain language.
2. The top 3 things to watch closely next month.
3. The 3 questions the numbers raise without answering.
4. No invented causes: the hypotheses stay as questions.

Real example: monthly summary → shared read

Input: at Brightside Studio, an 8-person brand agency, founder Maya pastes the last two summaries. An excerpt of the figures given:

Revenue: £41,200 (last month: £46,800)
Fixed costs: £28,400 (flat)
End-of-month cash: £19,100 (last month: £22,600)

Output, the part that matters:

Revenue fell by £5,600 (12%) from last month, while fixed costs held flat at £28,400. End-of-month cash is down £3,500.

Questions to dig into:

1. Did the revenue drop come from a one-off project ending, or a recurring client reducing their retainer?
2. Cash fell more slowly than revenue: is that a billing lag that will catch up next month?
3. Will fixed costs hold if this drop repeats for a second month?

Notice what the AI did not do: it did not declare that revenue fell “because a client left.” Maya gave it no such cause, so the hypothesis surfaces as a question, not a conclusion. That is the prompt working: it lights up the numbers without inventing the story behind them.

This prompt is part of the library that ships inside the kit, with Playbook P10 that makes it the monthly ritual.