A Year of Closed-Loop Programmes: What We Got Wrong
Reading this because you are specifying, buying or selling containers? Send the details and a grader answers in writing, usually the same business day.
Closed loop works, and our three biggest assumptions were wrong: customers overestimate their turn rate, dedicated fleets cost far more than pooled ones, and the damage schedule has to be agreed in writing before the first delivery rather than after the first incident.
We started offering closed-loop programmes properly in early 2025 — we supply reconditioned units, the customer fills and ships them, we collect the empties, wash or rebottle, and return them. The customer pays per turn rather than buying containers.
It works. Diversion on those units is effectively 100% because the container never leaves our custody chain, the customer carries no capital and no disposal cost, and the economics are good for both sides.
Three things we were confidently wrong about.
Wrong thing one: customers overestimate their turn rate
A closed-loop quote is built on turns per year. Ask a customer how many times they fill a tote annually and the answer is usually optimistic by a wide margin — not dishonestly, but because people count intentions rather than history.
Across our first-year programmes, actual turns came in at roughly 60 to 70% of forecast. The gaps were mundane: seasonal shutdowns, a line down for a fortnight, a product change, a customer at the far end who sat on empties for six weeks.
That matters because per-turn pricing with a low turn rate is worse for the customer than buying. At four or more turns a year over three years, owning wins. We had several programmes where the honest advice, after six months of real data, was that they should buy.
So we changed the process. New programmes now start with a three-month measured period at a transitional rate, and we price the actual turn rate rather than the forecast. Two of our first-year customers moved to outright purchase as a result, and both are still customers.
Wrong thing two: dedicated fleets are far more expensive than they look
Some customers want their own serial numbers back — unit-level traceability, usually for audit reasons. We priced that at a modest premium on the assumption that it was a bookkeeping difference.
It is not. A dedicated fleet breaks batch flow at every stage:
- Wash cycles are batched by residue class. A dedicated fleet means part-loading the line or holding units until enough accumulate.
- Yard staging needs separate positions per customer rather than one pool, which is the scarce resource here.
- Collections cannot be consolidated with other customers on the same corridor as freely, because the return has to come back to a specific pool.
- A damaged unit cannot simply be swapped from stock; it has to be rebottled or replaced and re-serialised into that customer's set.
Our original premium was about $18 a unit. The real cost is closer to double that, and we have repriced it. We also now ask harder about why traceability is needed, because in roughly half of cases the audit requirement was for documented wash certificates per turn, which pooled units provide perfectly well.
Wrong thing three: we left the damage schedule until there was damage
This was the genuinely avoidable mistake and it caused the only bad conversations of the year.
In a leasing arrangement, normal wear is the lessor's problem and damage beyond wear is the lessee's. That sentence is easy to write and useless in practice, because the whole argument is about which side of the line a given unit falls on.
We started with a general wear clause and no schedule. Then a customer returned units with a destroyed valve boss on four of them — somebody had been using a pipe wrench — and we had an unpleasant exchange about whether that counted as wear. It plainly does not, and we had given ourselves no document to point at.
Every programme now has a priced schedule attached before the first delivery:
| Condition on return | Treatment |
|---|---|
| Cosmetic staining, scuffing, coating fade | Normal wear — no charge |
| Gasket, cap, valve seat wear | Normal wear — no charge |
| Crazing from the stored product | Normal wear — no charge, rebottled at our cost |
| Destroyed or scored discharge boss | Charged — bottle replacement |
| Bent lower cage horizontal | Charged at jig labour; or scrap value if unrecoverable |
| Cage racked beyond jig recovery | Charged — frame replacement |
| Freeze deformation from overfilling | Charged — bottle replacement |
| Contamination with an undeclared product | Charged — full unit, plus wash-loop remediation |
| Unit not returned | Charged at replacement value |
Nobody enjoys agreeing that document and every customer has been glad of it afterwards. Having it in advance converts a dispute into a line item.
What we got right
Two things, for balance.
One-way pooling. Recovering containers from the customer's customer rather than from the filler's site turned out to be the most popular feature we offer and we had almost not bothered building it. It removes the filler's return freight entirely and stops their customers accumulating a wall of empties with the filler's name on them — which is a sales problem as much as a logistics one.
The diversion statement. Quarterly, scoped to the customer's own serial numbers, with units received by grade, the split between returned to service, rebottled, fabricated and regrind, pounds of HDPE and steel kept in service, modelled CO₂e avoided with the method shown, and landfill tonnage of zero. It has won accounts we would not otherwise have been shortlisted for.
Who closed loop does not suit
- Anyone whose totes are static fixtures — a wash-water reservoir, a dust-control tank, a rain catchment bank. Leasing a permanent asset is a more expensive way to own it. Buy technical grade and be done.
- Anyone filling once a year and storing for eleven months. The turn rate makes it uneconomic for both sides.
- Anyone under roughly twenty units a month at a single site. The collection overhead does not spread far enough.
- Anyone who needs a currently valid UN rating on every turn. That is a retest regime, not a leasing programme.
986 words · published 2 June 2026 by Marguerite Okonkwo, Operations lead at IBC Tote Market LLC. We correct posts rather than quietly deleting them; if something here is wrong, tell us and we will say what changed.