Anniversary posts are usually a list of achievements. This is more useful and considerably easier to write, because I remember the mistakes far better than the wins.
One: one wash chemistry, for four years too long
From 2011 to 2015 we ran a single caustic line. It was well-run, correctly titrated and temperature controlled, and for alkaline-compatible residues it did an excellent job.
For sugars it caramelised the residue onto the wall. For resins and adhesives it did essentially nothing. We graded those units as reconditioned because they looked clean and passed a test, and some number of them went to customers carrying residue the wash had never reached.
Nobody complained, which is the part that still bothers me. The complaint arrives at a customer's plant two weeks later and nobody connects it back to the wash. We only found out when we started the 48-hour hold-and-smell protocol in 2015 and discovered what our own output smelled like.
The hot-water line should have gone in three years earlier than it did. It is now the first thing I ask a partner yard during an audit, and we have decertified two yards over it.
Two: we built capacity when we should have shortened hauls
Through 2016 and 2017 the growth plan was more throughput in Savannah. More wash bays, more staging, more staff. It worked operationally and it was strategically backwards.
Freight is 43% of our own carbon footprint and a large share of what a customer pays. Concentrating volume in one location meant longer average hauls in both directions. We were getting better at a thing that should have been getting smaller.
The partner yard model from 2019 cut freight miles per tote by 41%. It should have started in 2016 and I resisted it because I did not want to depend on other people's standards. That instinct was reasonable and the answer was an audit programme, not centralisation.
I spent three years optimising the length of a journey that should not have been that long.
Three: equipment before vocabulary
We bought wash lines, a press, a test rig and a granulator before we had written down precisely what each grade meant.
The result was that for years our own staff used "clean", "washed" and "reconditioned" interchangeably in conversation with customers, and the customer heard whichever one suited their expectation. Disputes followed, and they were our fault.
The grading language on our grades page took about two days to write properly. It should have been the first capital-free thing we did and instead it was roughly the eighth.
The general lesson, which I have since watched other people learn the same way: in a business where the product is a judgement about condition, the vocabulary is the product. The machinery only enforces it.
Four: we published numbers before we could evidence them
Around 2014 we started quoting a cumulative diversion figure in sales conversations. It was derived from unit counts and an industry average weight, and it was approximately right.
Approximately right is not the same as defensible. When a customer's sustainability team first asked for the methodology in 2018, we did not have one that would survive scrutiny, and I had to say so after we had been quoting the figure for four years.
Theo rebuilt the whole thing from scale tickets and asset records, which is why our current numbers are smaller than they were and why we now publish the 88 kg per unit that reconditioning costs us. Smaller and evidenced beats larger and approximate, and I learned that the uncomfortable way.
Five: a decade of not asking for affidavits
We did not systematically ask fillers for prior-contents affidavits until 2013, and we did not make it a standing request until considerably later.
Every food-grade unit we could have documented and did not became an industrial unit worth roughly half as much. More importantly, every filler we did not ask was a filler who never learned that signing a ten-minute document doubles what their empties are worth.
Several customers now sign affidavits on their outbound empties as policy because we eventually explained it. That conversation should have started a decade before it did.
The common thread
Looking at those five together, four of them share a structure: we invested in the physical thing before the informational thing, and the informational thing was cheaper and mattered more.
| What we built | What we should have built first | Relative cost |
|---|---|---|
| A second wash line (eventually) | A residue-class decision rule | Free |
| More capacity in Savannah | An audit standard for other people's yards | A few thousand |
| A granulator and a baler | Written grade definitions | Two days of writing |
| A cumulative diversion figure | A methodology that could be checked | A week of work |
| A food wash bay | An affidavit request in the intake process | One form |
The pattern
The right-hand column is almost free and the left-hand column is capital. We consistently did the expensive one first, because capital feels like progress and a written definition does not.
The one I still cannot decide about
Not having a telephone.
The reasoning holds up: quoting a used container needs five written facts, and on a phone call at least one of them routinely gets heard wrong. Our quotes are better for it and our error rate is low.
But I know we lose enquiries from people who want to talk to someone, and I have no way to count them. Every year I reconsider it, and every year I conclude that the written record is worth more than the enquiries I cannot measure. I am genuinely not certain that is right.
What we got right
Briefly, because this is not that kind of post: used-first sourcing as a written trading rule rather than a value, refusing to let grade travel upward, buying the equipment that let us trace material, and never taking outside capital — which is the only reason the first three survived contact with a bad quarter.
Written by
Marcus Vreeland
Founder
Started the company in 2009 with a 1998 flatbed and a hand-written ledger. Still reads every decommissioning survey.