Closed loops only work if somebody owns the retrieval

Ship a container out one-way and there is a good chance you never see it again. At the receiving site it is emptied and becomes somebody else's problem — somebody with no incentive to return it. It gets repurposed behind their building, sold to a broker, or paid to be hauled away as waste.

Meanwhile you buy replacements. Each replacement is a new container carrying roughly 187 pounds of embodied CO2e, 42 pounds of virgin resin and 88 pounds of new steel. The container you already owned is sitting two hundred miles away doing nothing.

Why programmes fail

1. 01Retrieval is nobody's job. It is assigned to logistics, who are measured on outbound cost, and returns make their number worse.
2. 02There is no register. Without serial-level tracking nobody knows what is out, so nobody knows what is missing until the purchase order lands.
3. 03The economics are assessed one trip at a time. A retrieval leg looks expensive next to a single container. It looks trivial next to twelve containers over four years.
4. 04Nobody asked the receiving site. Customers will usually hold empties for collection if you tell them you are coming and give them somewhere to put them.

What actually makes it work

A register, a route and an owner. That is the whole design.

The register is serial-level: what you have, where it went, when it is due back, what testing is due on it. The route is scheduled rather than reactive, consolidating retrieval across several receiving sites so no leg is dedicated to a handful of units. The owner is one person or one supplier with retrieval in their scope, not a shared responsibility.

Where the maths starts working

What retrieval actually costs

Less than people assume, because empty containers consolidate. A retrieval leg picking up from four sites on an existing lane costs a fraction of four dedicated trips, and it is the same trailer that delivered outbound.

The comparison that matters is not retrieval cost against container cost. It is retrieval cost against replacement cost multiplied by the number of cycles the container would have delivered. A container that lives eight cycles instead of one changes the arithmetic by nearly an order of magnitude.

The reporting side

Programme customers get quarterly serial-level reporting: turns per container, wash cycles, repairs, recertifications, losses and a reuse-versus-reclaim split. That last figure is what a sustainability team actually needs, and it is impossible to produce without the register.

Which is the quiet argument for a programme even when the freight maths is marginal: you cannot report on circularity you are not tracking.

Getting the receiving site to cooperate

The site holding your empty containers has no incentive to store them and every incentive to make them disappear. Programmes that work make holding easy rather than mandatory.

• Tell them when you are coming. A site will hold containers for a known collection date and will not hold them indefinitely.
• Give them somewhere to put them. A marked area, even a painted rectangle, dramatically improves return rates.
• Collect on a rhythm they can predict. Monthly beats "when we have enough".
• Make the empties worth something to them, even nominally. A credit against the next order changes behaviour more than a clause does.
• Never ask them to clean or sort. That is what makes people stop participating.

What to track, and what not to bother with

The number that decides whether it works

Turns per container per year. Everything else is detail.

A container achieving four turns a year for six years has delivered twenty-four fills for one purchase. The same container shipped one-way has delivered one. That ratio is the entire business case, and it dwarfs the retrieval cost by a margin that makes precise freight accounting almost irrelevant.

When we tell customers not to do it

Scattered single-drop lanes, fleets under about a hundred units, or containers that regulation requires to be single-use. In those cases the honest recommendation is to sell the empties at the receiving end and buy reconditioned at the filling end, which achieves most of the material benefit without the retrieval overhead.

We would rather say that than sign somebody up to a programme that quietly stops in eighteen months.

Questions about programmes

How long does one take to set up? An audit of existing stock and a register can be done in a few weeks. The retrieval rhythm takes a quarter or two to settle, because it depends on your receiving sites developing a habit.

What happens when a container goes missing? It comes off the register and onto the loss figure, which is reported. Losses are normal at a low rate; what matters is whether the rate is measured and whether it is falling.

Can we run this ourselves? Certainly, and plenty of companies do. The parts that are hard to do in-house are the wash line, the testing and the consolidation of retrieval routes, which is where a supplier adds most of the value.

Does it work across borders? It becomes considerably more complicated — customs, wood-packaging rules, differing regulatory regimes. Domestic loops are straightforward and international ones are a project.