Kill the monthly report

3 October 20266 min readD6 Sigma

The monthly ritual

In many organisations the month ends and the reporting begins: several people spend days collecting numbers from different systems, pasting them into a pack of dozens of pages, and checking that the totals agree. By the time it is issued, the numbers are weeks old, and the meeting that discusses them spends its first half arguing about whose figure is right.

The answer is not a better template. It is to report less, define it once and let the machines do the collecting.

Seven steps

1. List every recurring report. For each one record the owner, the audience, the frequency, the hours it takes to produce and the last time anyone made a decision because of it. The last column is usually short.
2. Ask what decision it supports. If nobody can name one, stop the report. If two reports support the same decision, merge them. If the decision is made weekly, do not report monthly.
3. Define eight to twelve KPIs. For each: the definition, the formula, the source system, the owner, the target and the threshold that triggers action. Write it on one page and have the leadership sign it off.
4. Connect the data once. Identify the sources — the ERP, the scheduling tool, the spreadsheets, the forms — fix their quality and name a data owner for each. Nobody should retype a number that a system already holds.
5. Automate the refresh. Build the dashboard and the standard pack in Power BI or Excel so that they update on a schedule. Keep people for the commentary: what happened, why, and what we will do.
6. Change the meeting. The pack is read before the meeting. The meeting itself covers exceptions and decisions, and the decisions are logged.
7. Measure the saving. Count hours and steps before and after, and review the list of reports every quarter so that it does not grow back.

A worked example (illustrative)

Suppose four people each spend twelve hours a month assembling the monthly pack. That is 48 hours a month, or 576 hours a year — about fourteen working weeks at 40 hours. If stopping two reports and automating the data refresh removes 80% of that effort, about 460 hours a year come back, and the pack arrives days earlier.

The figures are an arithmetic illustration, not a benchmark. Your own baseline — measured before you change anything — is the only number that matters.

Pitfalls to avoid

• Automating a bad report. A faster report that nobody reads is still a waste. Remove first, automate second.
• KPI sprawl. If a dashboard has forty indicators, it has none. Keep one page and a drill-down.
• No data owner. Dashboards that nobody maintains decay in a quarter. Name an owner for every source and every KPI.
• Tool first. Many organisations already own Power BI or Microsoft 365. Decide what to report before you decide what to buy.