The impressive operation that shouldn't have existed

A fulfillment network I worked in had stood up an internal air-freight operation — moving its own inventory by air, between its buildings and its customers, to hit tight delivery promises. The idea had real merit on paper: air was faster, customers got their orders sooner, and demand for speed was growing. So the operation invested, and it kept investing. Over several years it got genuinely sophisticated: purpose-built loading equipment, a dedicated FAA-certified role managing the handoff between normal inventory and air freight, refined processes at every step. By any measure of operational craftsmanship, it was impressive.

There was just one problem nobody could answer: was it actually working? Air freight was a unicorn process — it didn't look like anything else the operation did, so nobody really knew how to measure whether it was succeeding. It ran for years in that fog. Walk past it on the floor and it looked half-dead. The data said the same: single-digit percentage of total volume. But it kept getting refined, because refining it was the thing everyone knew how to do.

I got asked to evaluate the operation end to end — including the part everyone had stopped looking at: what happened after the freight left the building. It turned out to be undramatic. A short drive to a nearby airport, a small crew loading a plane. It worked. It was also staggeringly expensive: tens of thousands of dollars a week in labor and support, sized for high, round-the-clock volume that no longer existed, propping up a process running at a fraction of capacity.

But the expense was the symptom, not the disease. Two deeper problems sat underneath it. First, it was over-engineered — built and staffed for a scale it never actually operated at, so most of that sophisticated support sat idle. Second, and more important, it was a band-aid on a bigger problem. Ask why freight needed to move by air at all, and the honest answer was: because product wasn't positioned close enough to customers to arrive on time by normal means. The air-freight operation wasn't solving a shipping problem. It was compensating, expensively, for a product-placement problem upstream.

I recommended winding the program down — and that's what happened. As the underlying product-placement issues got fixed, the need for air freight shrank toward nothing on its own, exactly as the diagnosis predicted. The expensive, elaborate operation that had consumed years of engineering attention simply stopped being necessary. It no longer exists.