The forest and the trees.

On a Sunday in February 1904, a fire started in a basement in downtown Baltimore. Within a few hours, engine companies from Washington had arrived by train to help. They knew how to fight a fire. They ran their hoses to Baltimore's hydrants and could not connect, because the threads were different. Companies came from Philadelphia, New York, Wilmington, and five other towns. Some of their hoses fit. Others did not. The fire burned for thirty hours and took 1,526 buildings across seventy blocks.

Nobody who came lacked skill. What some of them lacked was a hose that fit.

Most companies with fifty facilities have their own version of Baltimore. It burns slower.

The tree

Cargo theft is the problem everybody can see right now, and it deserves the attention. It is pure waste: freight you paid to make, gone, with nothing to show for it.

So size it for one company. The American Transportation Research Institute (ATRI) studied a shipper with more than 80 locations in the U.S. In the year before it changed anything, it had over 13 thefts. A standard trailer of its freight was worth $12,000 to $20,000. Scale that, as an illustration, to a company with 50 facilities and you get about eight thefts a year, or roughly $100,000 to $160,000 in stolen freight.

That is real money. If theft is the line you own, buying something to stop it is a reasonable call. I would ask what else that purchase costs you, because that is the tree. Here is the forest.

The bill nobody puts next to it

A company with 50 facilities is usually already running at least three main systems: a yard management system (YMS), a warehouse management system (WMS), and a transportation management system (TMS). Oftentimes there are multiple iterations of each of these systems, depending on the faciilty. There is also often already be a security product with its own infrastructure deployed at each site. That is a lot of IT complexity before you even start considering the purchase of yet another system that has the explicit goal of theft prevention. Buy a gate security product for every site and that adds to the pile.

The new product does what it promised. Then it has to tell the YMS what it saw, and the YMS has to tell the dock office, and the dock office has to tell everybody else. The TMS and WMS may also be interested in this thread. When the threads do not match, people become the coupling. Somebody at the guard shack keys the truck into a second screen. Somebody in the dock office radios the spotter to find out where the trailer went. Somebody walks the lot with a clipboard because two systems disagree about what is out there. None of it shows up on an invoice, so nobody adds it up.

So let's add it up. These are the assumptions in our own ROI model. Swap in yours:

Fifty facilities: two people in each dock office on a shift. That is 100 people.

The driver process: about 35 percent of their day goes to it. Checking trucks in, chasing paperwork, keying and re-keying, answering the radio. That is 35 people's worth of time.

The time you get back: the model assumes about half of that can be handed back once the whole visit lives on one record. That is 17.5 people's worth of time.

The cost: at $60,000 a year each, fully loaded, that is $1.05 million a year, modeled. Run two shifts and it is $2.1 million. My own guess is it takes another one or two people to keep four systems talking to each other, on top of that.

The $60,000 is on the low side. The median shipping and receiving clerk in this country earns $45,260 a year, according to the Bureau of Labor Statistics (BLS), and across private industry, wages average 70 percent of what an employer actually pays. That clerk costs closer to $65,000.

Now the other side of the ledger. On one line, roughly $100,000 to $160,000 a year in stolen freight. On the other, $1 million to $2 million a year in people's time spent holding your systems together. Theft has a long tail, too: ATRI, drawing on older RAND research, puts the indirect costs of theft, the delays and the downtime, at three to six times the value of the freight. Add all of that and theft at a company this size runs somewhere around $400,000 to $1.1 million a year. In the middle of that range, the people line is still bigger on one shift and more than twice as big on two. At the very top, it is about even with one shift. And if you ship premium freight, your theft line is bigger than mine, so run your own.

People are not waste

I want to be careful here, because I am not saying theft does not matter. Theft is pure waste. The time those people spend re-keying trucks is not. They are some of the most capable people in your operation, and right now you are spending them on clipboards and second screens.

Nobody has to go anywhere. Jim from security gets back to securing the site. The dock office gets back to running the dock. Give people their time back and they will find better things to do with it than re-keying a truck that already checked in.

A unilateral focus on theft, whether you are the company selling the fix or the company buying it, misses that. You can spend the year stopping $100,000 in theft and keep paying $1 million to hold the systems together.

Both at once

Go back to the shipper ATRI studied. Its fix was a check, run before the trailer moved. At certain locations, no shipment left without a pre-clearance that verified some combination of the driver's commercial license, the tractor and trailer numbers, the USDOT number, and the carrier's name and contact details. Thefts fell from over 13 to four.

Look at what that check asks: who are you, who do you haul for, and what are you here to take. Those are check-in and check-out questions. They are the questions a check at the gate is built to answer, and they matter: of the 677 cargo thefts Verisk CargoNet counted in the second quarter of this year, 158 were fictitious pickups, a truck showing up for freight that was never its own.

Ask those questions on a clipboard and it is one more job for the same people. Ask them once, in the same record as the check-in, the bill of lading (BOL), the move to the door, and the check-out, and nobody keys them twice. The driver checks in from his phone, gets loaded, and gets back on the road. The dock office already knows who he is.

That is what end to end can buy you: one record for the whole visit instead of four that people have to reconcile. Done that way, the theft check and the time you get back come out of the same work.

In other words, ensure that the forest isn't being missed for the trees in that solving for one painful variable (such as cargo theft) doesn't result in proliferating system complexity, causing hidden costs. You may even find some hidden opportunities (another tree for the forest), such as enterprise-level data and process standardization. We think you can kill several birds (and potentially plant a tree) with one stone, including reducing system complexity, which is what Primo Brands did.

One site, the whole visit

Last week I said start with one site. That has not changed. End to end does not mean all fifty facilities at once. It means the whole visit at the first site, from the gate to the dock and back out, on one record. Buy a tree and you buy it again at the next facility, along with another coupling and the people to hold it. Start with the whole visit and the second site runs the same record as the first.

So put your two numbers side by side. What did theft cost you last year? And how many people's worth of time across your facilities goes to holding your systems together? Reply with both. If the second one is smaller, I want to hear about it.

Jake Koppinger, Co-Founder and CEO, YardFlow by FreightRoll