Month-end variance commentary from budget versus actual

Draft month-end variance commentary from this budget versus actual data.

Period: {{month / quarter}}
Budget versus actual by line: {{paste the table — line item, budget, actual, or a CSV}}
Prior period actuals, for trend: {{paste, or "not available"}}
Materiality threshold: {{e.g. 7% AND $25,000 — both must be met}}
What I already know drove specific lines: {{paste anything you know — a delayed hire, a renegotiated contract, a one-off invoice. Leave blank if nothing.}}
Audience: {{e.g. department heads / CFO / board}}

Produce this.

**The filtered variance table.** Every line, with dollar variance, percent variance, favorable or unfavorable, and whether it clears the threshold. Apply the percent and the dollar floor together, so a small percentage on a large line still surfaces and a large percentage on a trivial line does not. Everything below threshold goes into a single "immaterial, no commentary" row with the net total, not a separate paragraph each.

**Commentary per material line.** Three sentences maximum per line: what happened, the most likely driver, and what it means going forward. Attribute each variance to a driver category and name it explicitly — volume, price or rate, mix, or timing. Where the data cannot distinguish between two drivers, say which two and what field would separate them, rather than picking the more flattering one.

**Timing separated from structural.** A dedicated section for variances that are purely when-it-landed: an invoice booked in the wrong period, a payment that slipped a month, a hire starting three weeks late. Give the expected reversal period for each. These must not be mixed into the structural list, because a timing variance treated as a trend produces a reforecast that is wrong twice.

**Internal versus external cause, and the resulting ask.** For each structural variance, say whether the cause sits inside the business (forecasting error, process, unplanned spend) or outside it (market rates, demand, supplier pricing). Internal causes get a named owner and a specific corrective action. External causes get a reforecast recommendation with the line and the revised assumption.

**Full-year implication.** For each structural variance, the run-rate effect if it continues unchanged for the rest of the year, stated as a number.

**What I should verify before this is sent.** Every figure you inferred rather than read directly from my data, and every driver attribution that rests on my note rather than the numbers. List them plainly so I can check them.

Do not restate the number in words when the table already shows it. "Marketing was $40k over budget, a variance of 18%" adds nothing next to the row it sits beside. The commentary earns its place by explaining the cause and the consequence.

Tip: Keep client or company identifiers out of the paste if you are using a consumer chat tool. Aggregate line items are usually enough for the commentary.