The Fazenda Santa Luzia occupies eight hundred and twenty hectares of rolling red earth on the border of São Paulo and Minas Gerais. There are four hundred and ten thousand coffee trees on the property. At harvest, a single mechanized picker can strip twelve tons of cherry from twelve hundred trees in an afternoon. The farm’s own dry mill processes between four hundred and six hundred bags a day during peak season. The office, in a converted stable, contains two desktop computers, a humidity chamber, and a cupping bench with twenty-four stations.
To someone who has spent time on small farms in Ethiopia or Rwanda, the first impression of a Mogiana fazenda is of a different industry entirely: one in which the unit of attention is not the tree but the block, not the lot but the silo, not the cherry but the ton. It is easy to assume, from this scale, that the attention itself is absent. It is not.
“Scale is not the opposite of care,” says Lucia Mazzaferro, whose family has owned Santa Luzia for four generations. “It is a different problem of care. Americans keep getting this wrong.”
What the picker does and does not do
Brazilian coffee is almost uniformly mechanically harvested — this is, to begin with, the most common criticism leveled at it in the specialty trade. A mechanical picker shakes a tree; the cherries fall onto a belt; the belt moves the cherries into a hopper; the hopper empties into a truck. The process does not distinguish between a ripe cherry and an under-ripe one. It does not, in theory, permit selection.
In practice, the modern picker — a French-Brazilian machine called the Jacto 5000, produced in a factory ninety kilometers from Santa Luzia — permits more selection than its reputation suggests. The operator adjusts the vibration frequency to favor cherries above a specific ripeness density. Under-ripe cherries, being firmer, resist separation. Over-ripe cherries, being softer, also resist in the other direction. The ripe middle falls. The pass is repeated, sometimes four times through the same block across a harvest, each pass calibrated for the ripening curve the block is on.
What leaves the field is, at a well-run farm, between sixty and seventy-five percent ripe cherry on the first pass and above ninety percent by the final pass. This is less selective than the obsessive multi-pass hand harvesting of Ethiopia or Panama. It is not, as the specialty trade often implies, no selection at all.
“The machine is a tool,” Lucia says. “A bad picker with a good machine is still bad. A good picker with a good machine is extraordinary. We have spent thirty years becoming good pickers.”
The honey process and its patience
Santa Luzia has, for the past fifteen years, produced an increasing portion of its harvest using a variation of the pulped natural process — known colloquially in Brazil as cereja descascado and internationally as honey. The cherries are depulped soon after harvest; the mucilage is left on the bean; the beans are dried on concrete patios under partial shade for twelve to eighteen days, raked every forty minutes by hand during daylight hours.
The process is labor-intensive in a way that the mechanical harvest is not. It requires judgment at every stage. It also produces a coffee with a significantly different character from the fully washed Brazilian coffees most Americans have in their heads: more red fruit, more overt sweetness, a fuller body, and the chocolate-and-hazelnut finish that has become, for better or worse, the international signature of Brazilian specialty.
On the cupping bench in the converted stable, at the end of the second day of my visit, Lucia pulls twelve samples from the current harvest and lines them up in numbered bowls. She tells me the block each came from. She tells me the pass. She tells me the dry-mill date. Then she tells me which one she is most interested in, which one she is slightly worried about, and which one she expects to change meaningfully over the next ten days as the oxidation curves settle.
The samples are not identical. They are not remotely identical. One of them — block 14B, second pass, pulped natural, dried sixteen days — is among the most purely pleasurable cups of coffee I drink all year: clean red apple, a honey sweetness, a long caramel finish, and a body that coats the tongue like silk. It is not a trophy coffee. It is not expensive. It is, in every reasonable sense, simply a superb cup of coffee, produced at scale, by a system that most of the specialty industry has trained itself to talk about with condescension.
The language problem
The word commodity is used in two different ways in the coffee trade, and the distinction matters. In the narrow sense, it means coffee traded on the C-market — a single standardized futures contract in New York, priced per pound, of a quality defined by the International Coffee Organization at a level most specialty coffee clears easily. In the broader sense, as used colloquially, it means coffee that is undifferentiated, interchangeable, and unworthy of attention.
A large Brazilian farm sells some of its coffee at the narrow commodity price. It does not follow that any of its coffee is, in the broader sense, a commodity. Santa Luzia, for instance, sells about a third of its harvest on the C-market, a third to large roasters at a specialty differential, and a third as single-lot microlots to individual roasters in Japan, Scandinavia, and Australia who pay, by any historical measure, significant premiums.
The first third funds the infrastructure. The second third funds the selection. The third third is where the reputation is made. All three come, in the end, from the same hills, picked by the same machines, processed by the same crew, with the same attention.
“Call us commodity if you want,” Lucia says, shrugging. “We will call you our customer.” She smiles. “The coffee is good. Tomorrow I will roast you the block 14B and you will think about this conversation again.”
She does. I do.