Before You Fix Your Department, Find Out What You're Actually Running
Ask three people at different levels of your department to name its top three priorities. Three different answers? You don't have an execution problem — you have an assessment problem.
The instinct in a new role is to act fast: reorganize, launch, show energy. But effectiveness and efficiency both start in the same unglamorous place: an honest, evidence-based look at what you're actually running. We rebuilt our unit assessment method on the Baldrige Excellence Framework — 48 diagnostic lenses across seven domains — and turned it into a working instrument. Six lessons survived the rebuild, and one was added by a reader.
1. Evidence beats impression. Everything in a unit exists four times: as designed, as documented, as practiced, and as it actually results. The findings live in the gaps. Don't ask whether a process exists — ask to be shown it working last month. A contradiction between the versions is evidence, not noise to reconcile away.
2. Unanchored ratings are opinions wearing numbers. A "3 out of 5" means nothing until levels 1, 3, and 5 are concretely defined per lens. And the classic trap: a brilliant method used by one team isn't a capability — it's a pilot.
3. Never average away risk. A unit can score a healthy 3.8 while carrying one exposure that could end its mandate. Our priority index rolls domains up by the maximum, never the mean — and a half-completed rating reads as unknown, never as low priority. A known severe gap must be structurally unable to vanish because someone missed a dropdown.
4. Assess the organization you have, not the chart. Target, funded, filled, and actually-available are four different headcounts, and demand is a fifth line the other four must meet. Then map who owns the cross-cutting work — compliance, quality, safety, data — in fractional FTE. In nearly every unit, one duty is owned by nobody and another hangs on a single person.
5. Count all the work — and all the money. Every unit runs an invisible second business: reports owed upward, policies to review, regulator requests. It eats 20–30% of capacity and reappears as "overload." And count the money the same way: a budget inherited as last-year-plus-4% assesses nobody. Can anyone state what one unit of your core output costs — and whether the forecast produced a decision this quarter, or just commentary?
6. Give it a rhythm — and know your governance stack. Pair the periodic assessment with a weekly operating review. And in a multi-level organization — national policy above, regional resources between, local accountability below — write the levels into the charter. Half the real findings are gaps between levels: the exception authority nobody holds, the clock that runs locally while the decision travels upward.
The instrument now lives in our Second Read suite, alongside its clinical and economic siblings. It opens with a completed sample — a combined UM and care management department of a health plan — runs entirely in the browser, and stores nothing anywhere: the assessment lives in a file you keep.
The thread through it all: be curious before being certain. The best unit leaders aren't the ones with the best answers on day one — they're the ones who ask the best questions first.
How does your organization assess its units — a real examination, or a form that gets filled in?