A Shoelace and a Dewatering Pump Have the Same Problem

Last week I wrote that my wife Jamie staffed her business with a team of AI agents. A few people asked the obvious question: what do they actually do?

Here’s one of them at work this week.

Jamie sells elastic shoelaces across a dozen channels. For years, deciding what to stock and when to reorder ran on a tool she paid a few hundred dollars a month for. It never quite earned its keep: keeping it useful meant constantly tweaking and configuring it around the quirks of each channel, and it still made mistakes. So she pointed one of her agents at it with a blunt assignment: is this thing worth what we pay for it, and what should we do about it?

Judging that fairly means knowing what the tool should have been doing in the first place. To make it interesting I will use an analogy from a very different business.

I spent years in equipment rental. Rental teaches you one lesson fast: a dewatering pump earns money in exactly one state, on a job, moving water. Sitting in the yard, in the shop, riding on a truck between sites, it earns nothing and costs you the whole time. The entire business is a fight to keep each asset in its earning state, and to know, to the dollar, what it costs you when it isn’t.

Inventory is the same animal. A shoelace earns at one moment, the sale. In a container, on a shelf, stranded in the wrong warehouse, aging in a storage bin, it’s that pump sitting in the yard. Just smaller, and with a SKU number on it. Follow any unit from cash to stock and back to cash, and you can see exactly where it stalls.

Which is what the agent found once it could see the whole cycle. A bestseller marked “plenty in stock” that was stranded in the wrong region, quietly selling out where the customers actually were. Product aging into storage fees on one end while the same item ran dry on the other.

A rental fleet and an online catalog look nothing alike, and they run on the same physics. Every asset is either earning or waiting. The whole job is to shrink the waiting, and to put a number on what the waiting costs.

That’s what the agent did this week. It made the nuances that matter visible: the whole cycle in view, the forecasts sharper, and stock moving at the pace the business actually needs. The tool it was sent to judge didn’t survive the review, and a few hundred dollars a month went back to the bottom line.

Often, the cheapest money in a business is the leak a fresh set of eyes finally catches.