Ask a small roaster what their green costs and they will quote the price per kilo they were offered. It is the number on the sheet, it is the number they negotiated, and it is the number that goes into the costing spreadsheet.
It is also, for a business buying in small quantities, a long way from what the coffee actually costs by the time it is in the hopper. The gap between the two is one of the least understood parts of this business, and it is wide enough to explain a large share of the difference between operators who are comfortable and operators who are not.
What sits between the offer and the hopper
The differential. Specialty coffee is priced as a differential over the exchange, not as a flat number, and the differential is where the negotiation actually happens. A buyer taking a container a month is quoted differently from a buyer taking four bags, on the same coffee, on the same day. This is not sharp practice. Small lots cost the seller more to handle per kilo and carry more risk of not moving.
Freight, and whether you are paying list. This is the one that surprises people most. A full container has a rate. A part load has a different rate per kilo, materially worse, and it also has more handling, more consolidation waiting time and more opportunity for something to go wrong. The small buyer pays list because there is nothing to negotiate with. The large buyer has an annual conversation with a forwarder and pays something else entirely.
Demurrage and storage. Free time at port is finite. If clearance takes longer than expected — paperwork, inspection, a public holiday landing badly — charges start, and they are per day and not small. A business that clears a container a month has a process and a relationship. A business that clears one a quarter is doing it from memory each time.
Financing the gap. Green is paid for early and sold late. Whether or not you have a facility, you are funding that gap: either you are paying for the money, or you are not earning anything on cash that is sitting in a warehouse as beans. That cost is real and almost never appears in anyone's cost per kilo.
Shrinkage and roast loss. Between what is invoiced and what leaves the drum there is moisture loss, chaff, sampling, and the batches that did not go to plan. The percentage varies by coffee and by profile. Whatever it is for you, it applies to the fully landed cost — not to the offer price — which means every cost above gets amplified by it.
The coffees you cannot buy at all. The least visible cost. There are lots that will not be offered to a buyer taking two pallets a year, because the seller would rather place the whole lot with one account. Small buyers are not choosing from the same list, and they generally do not know what they are not being shown.
Add these together and the landed, usable cost of a kilo of green for a small buyer is substantially above the offer price — and the gap is proportionally largest for the smallest buyers, which is the opposite of what the offer sheet implies.
Why this compounds rather than evens out
Every item on that list improves with volume, and several improve faster than proportionally.
The differential improves because you are worth more to the seller. Freight improves in steps — the largest single improvement is the jump from part loads to full containers, and it happens once. Demurrage falls because frequency creates process. Financing improves because the balance sheet strengthens. Access improves because you are now on a list you were not on before.
So two roasteries buying the same coffee from the same origin can land it at meaningfully different costs, and the difference has nothing to do with negotiating skill. It is structural, it is a function of volume, and it is the reason the pricing pressure a small roaster feels from larger competitors is not imagined.
What a small operator can actually do
Not everything here is a function of scale, and the parts that are not are worth attention.
Know your real number. Build a landed cost per kilo that includes freight, clearance, financing and loss, and use it in your pricing rather than the offer price. Many operators are less profitable on some lines than they think, and the only way to know is to calculate it.
Buy on fewer, larger shipments where cash allows. This trades against working capital, so it is a genuine judgement rather than an obvious win, but the freight step change is large enough to be worth modelling.
Consolidate with someone. Two or three roasters sharing a container is common in mature markets and rare here. It requires trust and some coordination, and it moves you across the biggest single cost step on the list.
Be honest about the access problem. If you are not being offered the coffees you want, that is information about your position in the market, not about your relationships. Treat it as a constraint to be solved rather than a personal slight.
The conclusion, which is not comfortable
For most of the costs above, the answer is volume, and volume is precisely what a small roaster does not have. That is the structural fact of a fragmented industry and no amount of good buying overcomes it.
What good buying does is make sure you are not losing money you did not know you were losing — which is worth doing, and which starts with calculating the real number rather than the one on the sheet.