The Cherry Doesn't Lie

Coorg, Karnataka. Late January. Ten weeks before Germany.

Before it becomes a note on a cupping sheet, it’s just fruit. And fruit doesn’t care about your expectations.

The farm is eighteen acres. That sounds like something until you’re walking it at seven in the morning and realise eighteen acres is just enough to be exhausting and not enough to be efficient.

The owner has been growing coffee here his whole life. He doesn’t use the word specialty. He uses the word careful.

I was there for a few days, pulled in through a mutual connection to shadow a consultant before my roasting course started in April. He was a Q Processor, one of roughly two thousand certified in the world. He worked exclusively with small farms. Not because small farms were noble. Because they were the ones that needed the work.

He had no interest in my opinions. I had the sense to keep most of them to myself.

The first morning, he handed me a sack and told me to pick. Randomly. Don’t select. Don’t look for the red ones. Just pick what’s in front of you.

This is called fly picking. It is not the harvest. It is reconnaissance.

Ten people cover ten sections of the estate and bring back what they find. The result is intentionally mixed: ripe cherries, underripe ones, a few already beginning to desiccate on the branch. The variety is the point. You’re not collecting good coffee. You’re taking the temperature of the crop.

Most estates use Brix readings at this stage. A refractometer, a drop of pulp juice, a number. It feels like data. He called it useless. Not entirely wrong, sugar content matters, but a single number without context is just a number. Sugar levels shift by variety, microclimate, by which side of a hill you’re standing on. The reading tells you what’s in the fruit. It doesn’t tell you whether the fruit is ready.

What tells you whether the fruit is ready is what he called the maturity point.

We spread a hundred cherries on a flat surface. He placed seven reference cherries in a line, a physical scale running from fully green on the left to dried and shrunken on the right. Three positions in the middle represented the ideal range. Not a description of ideal. An actual object you could hold and compare against.

Then we sorted.

Green. Unripe. Semi-ripe. Ripe. Semi-over-ripe. Over-ripe. Dried.

Twenty minutes. What you end up with is a distribution, not a score, not a grade, a picture of where the crop actually sits, spread across a table, unambiguous.

For a specialty harvest, you want seventy to eighty of your hundred cherries in the ripe and usable range. Fewer than that, you wait. More dried cherries than expected, hard, hollow, rattling at the far end of the scale, and you’ve already waited too long.

I asked him what the dried cherries meant exactly.

He said: the tree was ready. Nobody came.

In specialty production, a dried cherry in your lot is a defect. In the cup it reads as flat, papery, hollow. No fermentation protocol rescues it. No roasting skill compensates. The problem was decided weeks earlier, on the tree, by the timing of a pick.

Coffee doesn’t begin in the cup. It begins here, in a hand, on a branch, on a morning when the decision gets made or doesn’t.

The exercise takes under thirty minutes. No equipment beyond a flat surface and the hundred cherries already in your sack. Done weekly in the lead-up to main harvest, it builds a picture of crop readiness that removes guesswork from a decision that costs real money when it goes wrong.

It also teaches the pickers something. When you can hold a semi-ripe cherry next to a ripe one, physically, in your hand, the abstraction disappears. You’re not following an instruction. You’re making an observation. The difference between those two things matters enormously when you’re paying someone by the day to be precise.

He’d been using this method on farms across India for years. Almost nobody else was.

I asked why.

He shrugged. People prefer the Brix meter. It looks more professional.

There’s a version of coffee education that moves very quickly toward equipment and away from the crop. The grinders, the brew ratios, the extraction yields. The vocabulary of the counter, applied upstream to things that happen in soil and shade and weather.

I had been living in that version.

Standing on that farm at eight in the morning, a hundred cherries sorted into seven piles, I understood something I’d been circling for a while.

The roaster doesn’t decide the coffee. The picker does. Everything after that is working with a decision that’s already been made.

In April I would be in Germany, learning to apply heat precisely to green beans. I would learn roast curves and development ratios and the chemistry of the drum. I would get better at it.

But the farm would still be here. Making its decisions without me.

Roaster’s Note

∙ Cherry ripeness at harvest sets the ceiling of what roasting can achieve, not the floor

∙ Dried cherries in the lot mean defect in the cup; no roast profile corrects a late pick

∙ The maturity point method gives producers a visual, replicable standard that Brix readings alone cannot offer

Practical takeaway: Before evaluating a green coffee for roasting, ask the producer what their fly-pick data looked like. If they don’t have any, that tells you something too.

— Roaster on the Road