The report that ate the first week of every month.

I have a weakness for the reports nobody wants to touch — the one that eats a week of every month, that everyone works around instead of fixing, because fixing it is never this month's problem. I can't leave them alone. This is the story of one of them.

It belonged to an online-sales division I was working with: the monthly performance scorecard.

The ritual

Ten exports came out of the CRM every month, one per report, and not one of them shared a key with the others. So every month, someone sat down and rebuilt the whole picture by hand — filtering each file to the region, cleaning it by its own particular rules, rolling it up division by division and person by person, until it became one scorecard an executive could use to actually coach a team.

It worked. It also quietly cost the first week of every month, and it got heavier each time — more months to compare, more people to track, more places for a number to drift. And since it was rebuilt from scratch every time, it was only ever as consistent as whoever happened to build it that month.

That last part is what I couldn't get past. It wasn't the tedium. It was that a real decision — about real people's performance, the kind that changes how someone gets coached — was resting on a spreadsheet stitched together from nothing, thirty days at a time.

I didn't want to make it faster. I wanted to make it stop.

There's an easy version of this job: make the manual process a little quicker. A cleaner template, a few macros, done. I've watched a lot of “sales ops” end right there — and it's still a spreadsheet, still one bad copy-paste from being wrong, still somebody's week.

So I built the actual thing. You drop the ten exports in. It cleans each one by its rules, rolls everything up, and rebuilds the entire scorecard — the leaderboards, the driver correlations, the flags for who's slipping and who's clearing benchmark. The second a comparison month exists, the month-over-month, quarter, and year views show up on their own, pooled from the raw counts instead of averaged, so the trend is real and not a smear.

The dashboard was never the point. The point was that the report stopped being a project and became a drop of the files.

What changed

The week came back. The executive got the same scorecard, built the same way, every month — and the trends and coaching flags that used to take real effort, or didn't exist at all, were just there, every time, without anyone having to find time for them.

But the part I'm proudest of is boring: you can trust it. Every number traces back to the export it came from. The definitions are written down. The same person is counted the same way in June as in September. That's the whole difference between a report that looks right and one you can defend out loud when someone challenges it — and someone always challenges it.

Why this is the work I do now

I've come to think this is what operations actually is. Not one big transformation — a hundred small, specific, unglamorous processes that are too particular to buy off a shelf and too important to drop, each one quietly eating hours because nobody has the hours to spare to fix why it eats hours.

That's the work I take on now. So if there's a report in your world like this one — the one that eats the first week of every month — that's exactly the kind of thing I want to hear about.