Planned versus reactive: the ratio that runs your operation

Every operation splits its work into two piles: things it decided to do, and things that happened to it. Maintenance you scheduled versus breakdowns you attended. Orders you planned for versus expedites you absorbed. Hiring you sequenced versus the panic backfill after someone quit.

The ratio between those piles is the most honest health metric a business has, and almost nobody can state theirs out loud. Ask a room of managers what percentage of last quarter's work was planned, and you'll get a pause, then a guess, then a debate about definitions. That pause is the finding.

Reactive work costs a multiple of the same work planned — not just in the emergency premium you pay, but in the downtime around it, the commitments it breaks, and the attention it steals from people who were meant to be doing something else. A single unplanned failure quietly bills three departments.

And the drift is invisible month to month. Nobody wakes up to a memo saying "we became a reactive operation this quarter." It arrives one exception at a time, each perfectly defensible, until firefighting is the culture and the calendar exists mainly to be interrupted.

Why it stays invisible

Because the evidence is scattered across documents nobody reads line by line. Invoices, work orders, tickets, expedite approvals — each one is a single event in somebody's inbox, and the pattern only exists in aggregate. A veteran might feel the drift; they can't prove it, and what can't be proven doesn't get budget.

This is the most under-appreciated thing intelligence does for an operation, and it's nothing like the demo. Read every one of those documents at line level. Resolve each line to the actual asset, vendor, customer, or process it belongs to. Baseline what normal looks like for that thing. Then the ratio stops being a feeling and becomes a number you can watch move — per asset, per line, per month.

What you do with the number

You stop arguing about whether things are getting worse and start seeing which specific things are. The repeat repair that shows up three times in five weeks across two vendors — invisible on any single invoice, obvious once the lines resolve to one asset. The customer whose order cadence just broke a twelve-month pattern. The process step that generates a disproportionate share of your exceptions.

Then you act on the schedule instead of the emergency. Not because a model predicted the future perfectly, but because what was and what is, read together and honestly, tell you a great deal about what's going to be.

You don't need a crystal ball to stop being surprised. You need your own history, read line by line, by something that never gets bored.