This process uses the data dictionary, the flat file, and the monthly report template.
What you’ll end up with
- A correct set of monthly totals: raw numbers summed from the month’s weeks, every ratio recalculated from those totals.
- A three-to-five-page report, filed in the monthly report archive, explaining what drove the result, not just what it was.
Before you start
- Every week in the month has been loaded and checked through the weekly reporting process. A monthly rollup built on unchecked weeks inherits every one of their errors.
The process at a glance
1
Roll up the month
Sum the raw flat file inputs across the month’s weeks, then recalculate every ratio from the totals.
2
Check the rollup
Confirm the monthly totals reconcile against the individual weekly reports.
3
Analyze
Compare against target and prior periods, and work down the KPI tiers to explain what drove the result.
4
Write
Fill in the monthly report template, executive summary last.
5
Quality check and deliver
Have a second reader check it, then send with the headline and the ask.
6
File it
Log the report in the monthly report archive and register it if it’s a recurring cadence in the data dictionary.
1. Roll up the month
Filter the flat file to every week whose dates fall inside the month, add up the raw inputs (spend, revenue, orders, and so on) across those weeks, and only then recalculate every ratio, ROAS, conversion rate, contribution margin, from the summed totals. A spreadsheet pivot table does the summing in a few clicks; see the flat file template for the exact steps. If a month straddles a partial week at either end, either assign that week fully to the month it mostly falls in, or split the week’s rows proportionally, whichever your data dictionary specifies. Do it the same way every month.2. Check the rollup
Add the month’s four or five weekly totals for a single input, like ad spend, and confirm it matches the monthly sum. This lifetime-style check catches a missed or double-counted week before it reaches the report; if the numbers don’t reconcile, a week is either missing from the flat file or was pulled into the rollup twice.3. Analyze
Compare the month against its target, against last month, and against the same month last year if the business has real seasonality. Work down the KPI tiers: when a tier 1 outcome missed or beat its target, the tier 2 drivers underneath it say why, and tier 3 explains why a driver moved if that level of detail is needed. Check the change log for anything that happened during the month before concluding a cause.4. Write
Fill in the monthly report template. Write the body first and the executive summary last, since it’s hard to summarize an analysis that isn’t finished yet. Every number in the report should answer what happened, why, and what to do about it; a number that only states a fact is data, not reporting.5. Quality check and deliver
Have someone other than the writer read the report before it goes out. Ten minutes from a second reader catches more than an hour of re-reading your own work, and this is the cadence where an error does the most damage to credibility, since a monthly report usually reaches a wider or more senior audience than a weekly one. Send on the scheduled day with a short message stating the headline result and the ask.6. File it
Log the report in the monthly report archive, including its headline result, so the archive stays useful as a scannable history rather than a folder of files nobody reopens.Related resources
- Data quality and rollup concepts Why summing inputs and recalculating beats averaging.
- Weekly reporting process The process that produces the weeks this report rolls up.
- Monthly report template The fill-in template this process produces.
- Monthly report archive Where the finished report gets filed.
- Quarterly reporting process How three of these months become the next report up.