The slow part of the early coffee trade was not the sea. A ship under sail could work the Red Sea, round the Cape and reach a French port inside a few months, and while the weather made that number move, the crew broadly knew what they were looking at. The slow part happened at anchor. A ship sat off a mud-brick port on the Yemeni coast, waiting for enough coffee to exist in one place to be worth sailing home with.
Jean de la Roque's accounts of two French trading voyages, which Jonathan Morris reports in Coffee: A Global History, record that ships at Mocha in 1709 and again in 1711 took about six months to fill a single hold. A Dutch agent the French met at that port budgeted a full year for the same job. That is loading time, before the voyage home even starts. The constraint was not the ships and it was not demand. The world's entire commercial coffee crop came off terraced smallholdings in the Yemeni highlands, and it had to be assembled at a coastal market one camel load at a time before any of it could go anywhere.
Six months is not a colourful detail about the past. It is a number that tells you what the bottleneck was, and the bottleneck was never transport.
The route that metered the flow
Growers in the interior highlands sold down the mountain, and the whole trade funnelled through one inland town on the coastal plain, Bayt al-Faqih, where the crop changed hands and sat in warehouses. Animals took it the last leg, to the two shipping points the English wrote down as Mocha (Al-Makha) and Hudaydah. Morris adds the part that gets left out of the romantic telling: the money behind all of it was Gujarati. Banyan merchants held the Yemeni credit networks, which on his reading makes them the likeliest people to have got the cultivation started at all. The people financing the world's only coffee supply were Indian, not Arab and not European.
Set that route against the volume it was carrying and the six months stops being surprising. Wild Ethiopian coffee had the trade to itself until the 1540s, when planting crossed the water into the Yemeni interior. For the two hundred years that followed, nowhere else on earth grew coffee for sale. Morris puts the population living on those terraced uplands at around 1.5 million by the 1700s. Every cup drunk in Cairo, Istanbul, Isfahan, Venice, London and Amsterdam came off an area you could walk across in a week.
There was no buffer anywhere in the system. No warehouse held a year of stock, because there was no year of stock. The ship was the warehouse, and filling it meant waiting for the harvest, the terraces, the camel trains and the coastal market to deliver in the small increments they were capable of.
The scale, in numbers you can check
By the 1720s, Red Sea shipments were running at roughly 12,000 to 15,000 tonnes a year, a figure Morris reads as the whole world's supply for that period.
Convert that into the unit the modern trade uses and the comparison becomes uncomfortable. A trade bag is 60 kilograms. Twelve thousand tonnes is 200,000 bags. Fifteen thousand tonnes is 250,000 bags. The USDA's July 2026 world circular forecasts 2026/27 production at 189.667 million bags. So the entire planet's coffee supply in the 1720s was somewhere between one seven-hundred-and-sixtieth and one nine-hundred-and-fiftieth of what a single year now produces.
The lead time also did something to the coffee itself, and that part ended up in people's cups. The first sacks aboard sat in the hold half a year longer than the last, in a wooden ship, in Red Sea heat, before a voyage that added months more. Kenneth Davids sets out the direction green coffee travels as it ages in Home Coffee Roasting. Acidity and brightness drain away while the body thickens, and warm humid storage compresses that from a matter of years into a matter of months. A ship's hold is warm humid storage.
So the taste Europe learned to call coffee was an aged taste, low in acidity and heavy in body, and the trade eventually valued it on purpose. It had a name, sweating, and a market. It also had counterfeiters. William Ukers, writing from inside the American trade in 1922, records that steaming green beans to imitate the effect was eventually ruled adulteration under the 1906 Pure Food and Drugs Act, which is a sentence worth sitting with: there was enough money in a shipping accident to be worth forging. India's Monsooned Malabar, where green coffee is deliberately left open to the monsoon until the beans lighten and puff up, is the version that survived legally. All of which is a reason to know that green coffee does not keep forever, and that what ages it is heat and moisture rather than the calendar on its own.
What a six-month load actually cost the buyer
Time at anchor is capital sitting still. A merchant who bought coffee across six months in Mocha had financed the whole purchase before a single sack reached a buyer in Europe, and had carried the risk of price, spoilage, war and shipwreck for the entire period. That is the real reason the trade organised itself the way it did.
It is also why the British East India Company set up a permanent buying post at Mocha in 1618, more than three decades before anyone in Britain could buy a cup of it. You station a man at the source when you cannot simply arrive and purchase. The English were coffee traders for years before they were coffee drinkers, which is a better explanation of how the drink reached London than any story about a single coffee house. Morris notes that the clerks had not settled on a spelling either, entering the goods as cowa, cowhe, cohoo and coffa, which is what an unfamiliar commodity looks like in a ledger before it has a name.
What this will not do
Knowing the loading time at Mocha will not make you a better judge of a coffee. A port name is not a quality claim, and it never was one. The word began as a shipping label rather than a place of origin. When the Qasimi imams unified Yemen in 1638 and took Zeila with it, coffee from both sides of the Red Sea started leaving through the same harbour, so Ethiopian and Yemeni beans reached buyers under one name. Davids notes where that ends up: no coffee has left the actual harbour in more than a hundred years. The label outlived the port it was named for, first as a market term printed on sacks, later as a café word for coffee with chocolate in it.
This history also will not tell you how old the coffee on your counter is. Lead times explain why a supply chain behaves the way it does. They do not date a specific bag, and no amount of context substitutes for the information a roaster chooses to print. That is a separate problem, and it is worth knowing what a bag can and cannot tell you before you infer anything from a country name.
And it will not make the modern trade look fast by comparison in a way that flatters anybody. The trade got faster in exactly one place.
Which part is still slow
The assembly step is the part that changed. Washing stations, trucks, dry mills and containers collapsed the interval between a picked cherry and a loaded ship from months into weeks. That is a genuine and large improvement, and it is the reason a modern origin can respond to a buyer within a season rather than within a year.
The agriculture did not change at all. James Hoffmann's World Atlas of Coffee gives up to nine months between flowering and a cherry worth picking, and about three years between putting a tree in the ground and getting a real crop off it. Harvest is a season, not an event, and picking only ripe fruit means going back through the same trees several times. Those numbers are the same numbers la Roque's ships were waiting on, and no logistics improvement touches them, which is the whole reason coffee still has seasons and your favourite origin still disappears.
The narrow water did not go away either. In 2026 the International Coffee Organization reported the Strait of Hormuz effectively closed from 28 February, adding 10 to 14 days to Asia to Europe transit, with bunker fuel up 68 percent, container spot rates roughly doubled and fertilizer up 25 percent, before a gradual reopening on 22 and 23 June 2026. Hormuz is not the Red Sea. It is the entrance to the Persian Gulf, a different chokepoint on a different route, and the distinction matters because coffee writing has a habit of blurring the two. The transferable point is narrower and more durable: this commodity has always moved through a handful of tight passages, and the cost of one of them closing lands on the shelf months later.
Which is the practical version of the six-month figure. Lag is structural in coffee. It is why the price on your bag does not fall when the market falls, and why a roaster who has run out of something is usually telling you the truth rather than managing you.
We work from Kampala, which puts us at the assembly end of this rather than the retail end. The USDA's July 2026 circular forecasts Uganda's 2026/27 crop at 7.16 million bags, roughly 6.0 million of it robusta, and all of the export volume leaves as green bean. The step la Roque watched still exists here. Coffee is still gathered in small quantities from a great many growers and consolidated before it can be sold as a lot. The camels are trucks and the coastal market is a set of washing stations and a dry mill, and the arithmetic underneath is recognisably the same arithmetic.