Every Tuesday a number comes out of Wakulima House and travels fast. "AA fetched USD 343 a bag." By Thursday a farmer in Nyeri or Kirinyaga has seen that headline and done the arithmetic: 343 dollars is about 44,000 shillings, divided by 50 kilos is 880 shillings a kilo — so why does the factory notice say 117?
Both numbers are true. They just describe different things at different points in a chain. This piece walks that chain end to end.
Where the price is discovered
Almost all Kenyan coffee is priced at the Nairobi Coffee Exchange. The NCE has operated as a spot market since 1935 and handles over 90 percent of coffee sales, while around 550 cooperatives market more than 80 percent of the country's coffee.
The mechanics are deliberately slow and public. Auctions run every Tuesday unless there is insufficient coffee on offer. Around 9kg of each lot sits in the Nairobi Trade Sample Room the week before, divided into 250g parcels and sent to participating and prospective buyers. On the day, each lot is displayed with its outturn number assigned at the dry mill, its lot number, grade, bag count, total weight and the producer's name.
That sequence matters. Buyers have cupped your coffee before they bid on it. Price is a response to a cup, not a guess about a region.
How a buyer arrives at a number
A bid is built in layers.
The floor is the world price — ICE Arabica futures, the "C price." That is the buyer's opportunity cost of buying Kenyan rather than anything else.
On top of that sits the differential: what this specific lot is worth above or below the C. That is driven by cup score and profile, grade and screen size, defect count, moisture, lot size and consistency, certification, and the reputation of the factory and its outturn history. A clean, sweet, blackcurrant-forward AA from a factory that delivers the same thing every season is worth a premium a buyer can resell. A grassy, under-fermented lot of the same grade is not.
Then come the buyer's own economics — warehousing, finance cost, shipping schedules, and crucially whether they already have this coffee sold onward. A buyer with a roaster contract in hand bids differently from one buying on speculation.
The competitive layer matters too. In Sale 41, the top five buyers took 77 percent of the coffee on offer. When demand concentrates among a handful of dealers, the depth of bidding on any single lot thins out, and your lot's price depends heavily on whether two of them happen to want it.
What a reserve price is, and what it isn't
This is the most misunderstood instrument in the whole system.
Under the Nairobi Coffee Exchange Trading Rules, the reserve price is the minimum price set before the auction by a broker, a miller or a grower-miller in consultation with the grower. It is your floor — the price below which you would rather keep the coffee than sell it.
It is set by reference to the prevailing C price, the grade, and the assessed quality of the lot. But bids may simply not reach it. If a lot carries a reserve of 220 and bidding stalls at 160, the auctioneer may record it as a "noted bid," and the buyer who made that bid negotiates with the auctioneer after the auction — at which point a price below the reserve may in fact be accepted.
So a reserve price is a negotiating position, not a guarantee. Three things follow from that, and every cooperative committee should understand them:
Set it too low and you leave money on the floor. Buyers see the catalogue. A soft reserve signals a seller in a hurry.
Set it too high and your coffee doesn't clear. It goes back into the warehouse and onto a later catalogue — and every week it waits, it accrues storage, insurance and financing cost, and it ages. Withdrawn lots also carry a reputation cost; coffee that keeps reappearing starts to look like coffee nobody wants.
A reserve is only as good as the valuation behind it. This is where the grower's relationship with their broker is decided. If the reserve is set without genuine consultation — or set to suit the broker's volume targets rather than the lot's worth — the farmer carries the loss and never sees where it happened. Ask to see the reserve before the sale, not the result after it.
Turning a bag price into a cherry price
Now the arithmetic that causes all the confusion.
Auction prices are quoted per 50kg bag of clean (green) coffee. Sale 41 gives us a live example: 990,385 kg sold for Ksh755 million, at an average of Ksh38,118 per 50kg bag. That is Ksh762 per kilo of clean coffee.
Then apply outturn. Cherry is mostly water, skin and mucilage. Roughly 6.5 kilos of cherry yield one kilo of clean coffee, though the true figure varies by variety, ripeness, season and how well the factory pulps and dries. Divide Ksh762 by 6.5 and you get about Ksh117 per kilo of cherry — which is exactly the gross figure Sale 41 translated to, before deductions.
Then remember the grade split. A single delivery does not come out of the dry mill as AA. It comes out as AA, AB, PB, C, TT, T, UG1, UG2 and mbuni. In Sale 41, AA averaged USD 343.05 while UG2 averaged USD 158.23. Your payout is the weighted average across all of it. This is why quoting the AA headline to a farmer is misleading — almost nobody is paid the AA price, because almost nobody produces only AA. Better cherry selection shifts the mix upward, and that shift is worth more than most farmers realise.
The deduction stack
Between the auction hammer and the farmer's account, the money passes through costs. Broadly, in two tiers.
Chain-level: dry milling and grading, brokerage or marketing commission, warehousing and handling, exchange and settlement charges, statutory levies and county cess, insurance, and transport.
Cooperative and factory level — this is the tier farmers ask about most, and it is where real variation lives:
- Wet mill running costs: water, pulper power, repairs, effluent management
- Drying table labour, turning, skin-drying and conditioning
- Cherry clerks, weighing staff, and casual labour at peak
- Security services at the factory and store, which are not optional when parchment is sitting on site
- Factory manager and society staff salaries
- Committee allowances, AGM costs, audit fees
- County permits, licences and compliance costs
- Recovery of input advances — fertiliser, chemicals, seedlings issued on credit
- Loan servicing, and in many societies, historical debt
That last one deserves honesty. Two cooperatives with identical coffee and identical auction results can declare payouts 30 shillings apart, and the difference is usually not the coffee. It is overheads, staffing ratios, energy costs, and legacy debt.
The scale of the spread is visible in public numbers. In Kirinyaga's 2024/25 season, 27 of 75 wet mills declared rates between Sh100 and Sh145 per kilogramme of cherry, with Gacami Coffee Factory topping the list at Sh145.10. Of Sh7.4 billion received by cooperatives in that county, Sh6.99 billion reached farmers, with the balance covering factory operation costs. That is roughly a 5.5 percent deduction at the society level — good discipline. Many societies are nowhere near it.
What the new law changes
The framework has shifted substantially, and it is worth knowing what you are entitled to.
The Coffee Act 2023 replaces the old regime built on subsidiary regulations, entrenches the Direct Settlement System requiring proceeds to pass through a central settlement system before payment to growers within 14 days, caps cooperative deductions at 10 percent, and establishes a Price Stabilization Fund financed through the coffee levy.
At the June 2026 rollout in Kirinyaga, the government pushed further: at least 80 percent of proceeds paid directly to farmers with service providers sharing the remaining 20 percent, and payment within five days. Officials cited farmgate prices rising from about Sh50 per kilogramme two years ago to as high as Sh158 in some factories, with a stated target of Sh250.
The 10 percent statutory cap is the line farmers should hold their committees to. If your society is deducting more, it owes you an explanation in writing.
What actually moves your number
Strip away the noise and the levers are few:
Cherry selectivity. Picking only ripe cherry is the single cheapest quality intervention available and it moves your grade mix, which moves your weighted average.
Processing and drying discipline. Fermentation control and slow, even drying separate a cup that earns a differential from one that does not.
Outturn efficiency. Losses at the wet mill and dry mill come straight out of the farmer's kilo.
Cost discipline at the society. Every shilling of overhead is a shilling off the payout. Read the audited accounts.
Choice of broker, and involvement in the reserve. You are entitled to be consulted. Use it.
Timing into the sale. Which catalogue your coffee lands in, and against what competing volume, is a commercial decision — not an administrative one.