The situation
A large healthcare data-operations organization ran its project portfolio off a weekly executive rundown. The problem wasn’t a lack of reporting — it was that the dates weren’t true. Teams entered due dates by hand, let them slip quietly, and the schedule never showed the downstream effect. Every week the same projects promised the same near-term finishes, and every week the dates moved. The executive sponsor wanted to know when work would actually land, not watch a status loop reset itself.
What was broken
The team had a capable scheduling tool but wasn’t using predecessor logic — the dependency links that let a schedule compute a real end date instead of a typed-in one. That wasn’t an oversight. Turning the logic on would have forced parts of the portfolio red and exposed end dates the managers didn’t want exposed. The tool could already tell the truth. The process was arranged to avoid it.
What I did
I didn’t fight the existing process head-on — that would have produced a turf fight and no system. Over about sixty days I built a parallel cadence in a separate workspace while keeping the managers’ existing artifacts running, so nothing broke and no one had to defend their dates before there was something better to replace them with. I rebuilt the schedules so predecessor logic actually worked and the end dates computed themselves, and I stood up a dashboard that reported forecast dates rather than promised ones. Around week seven I began cutting projects over one at a time, retiring each old artifact as its project moved, until the new system was the one driving the weekly rundown.
The result
The rundown started showing forecast truth — when work would finish, not when someone hoped it would. Slippage became visible the week it happened instead of the week a deadline arrived. The cadence ran on the system rather than on manual effort: I automated roughly sixty percent of my own workweek into it. And it outlasted me — the person who took over the role kept running it after I moved on, because the leverage it created was structural, not personal to me.
Why this is the work I want to own
Most operations run the way that schedule did — on numbers that look fine and quietly aren’t. The hard part was never the software. It was being willing to expose the real picture, absorb the short stretch of a portfolio going red, and build the system that keeps the truth in front of the people making decisions. That’s the work I want to do on my own balance sheet rather than someone else’s.