Steeping the perfect read...
Quality takes time. One bean at a time.
ICB's cross-catalogue data shows one estate sold by 23 roasters as everything from light naturals to medium-dark anaerobics. The estate is provenance; the roaster is the author.
Ratnagiri Estate, one farm in Karnataka, appears on ICB under 23 different roasters. Across those roasters it is sold as 80 separate coffees, and those coffees cover every roast level from light to dark, nine different processing styles, and a price range of Rs 295 to Rs 2,200 per 250 grams. That is a 7.5-fold price spread, with a flavour spread to match, all carrying the same estate name.
Hold that for a moment, because it is the whole puzzle. If the farm is the constant, with the same soil, the same elevation, the same growing season, then something downstream is producing all of this variation. The estate name on the bag stays fixed while the coffee inside changes completely.
This is the kind of pattern that only shows up from a neutral vantage point. No single roaster publishes a comparison of its version of an estate against a competitor's. But a directory that carries every roaster's catalogue at once can line them up, and when you do, the roaster stops being invisible and starts looking like the largest variable in the cup.
An estate name on a bag is a sourcing claim, not a grade. Nothing formal governs the word "estate" in Indian coffee the way a GI tag protects a region name or an SCA cupping score grades a specific lot. It tells you which farm the green came from. That is genuinely useful information, but it is provenance, and buyers routinely read it as a flavour promise it was never making.
The reason one estate can become so many different coffees is that three separate variables sit between the farm and your cup, and they stack on top of each other.
The first is the estate itself. A farm like Ratnagiri does not sell one green coffee. It runs its own processing programme and sells naturals, anaerobics, carbonic macerations, double-ferments, and experimental lots as distinct products, often split further by varietal. The second is green selection: each roaster chooses which of those lots to buy, and they do not all choose the same one. The third is roast development. Having bought a lot, each roaster decides how light or dark to take it, which on its own can move a coffee from a fruit-forward cup to a chocolate one.
So this is not a story of identical green coffee turning out differently. It is a story about how much the single word "estate" hides. By the time a Ratnagiri coffee reaches you, the farm is the only thing two roasters' versions are guaranteed to share.
Info: This is measurable only because the comparison is neutral. A roaster has no reason to show how its version of an estate stacks against a competitor's, so the comparison never appears in roaster marketing. A directory that carries every roaster's catalogue can place them side by side, which is what makes the roaster effect visible in the first place.
Ratnagiri's 80 coffees span all five roast levels (light-medium 27, light 21, medium 14, medium-dark 11, dark 2) and nine processing styles, led by anaerobic (16), natural (15), experimental (13), and washed (11), with carbonic maceration, double-ferments, washed-naturals, and honey filling out the rest. The same estate is at once a light natural, a medium-dark anaerobic, a carbonic yeast maceration, and a barrel-aged lot, depending on which roaster you buy from.
The individual coffees make the spread concrete. Caarabi sells a light YC Culture Naturals at Rs 1,688. Quick Brown Fox lists a "Bioreactor Thermal Shock Naturals" lot, processed with lab-inoculated microbes, at Rs 1,249. Blue Tokai takes a YC lot to a light double-ferment at Rs 1,094. Subko roasts an SCA 86+ anaerobic as a light filter roast at Rs 995. Naivo goes the other way with a medium-dark anaerobic, Culture Naturals Rose, at Rs 845. Hill Groove sells a carbonic yeast maceration at Rs 799, and Fraction 9 lists a value-tier anaerobic at Rs 779. Same estate name on every one, and almost nothing else in common. That a single farm can support a lab-inoculated bioreactor lot and a value-tier anaerobic in the same season says a lot about how far processing experimentation in Indian specialty now reaches. The estate is functioning less like a fixed origin and more like a shared laboratory.
Same estate, different authors: Ratnagiri across roasters
The experimental, light end
- Roaster: Subko
- Lot: Project Pearl (SCA 86+)
- Roast: Light (filter)
- Process: Anaerobic
- Why it shows the effect: A light filter roast built to preserve the ferment character, Ratnagiri as a clean, high-clarity experimental cup.
The darker, fuller end
- Roaster: Naivo
- Lot: Culture Naturals Rose
- Roast: Medium-dark
- Process: Anaerobic
- Why it shows the effect: The same estate and a similar process taken to a medium-dark roast, giving a fuller, sweeter, lower-acid cup from the same farm.
Then there is the price. Across the 23 roasters, Ratnagiri ranges from Rs 295 to Rs 2,200 per 250 grams, with a median of Rs 890. A 7.5-fold spread from one estate is not a quality ladder. It reflects which lot was bought, how complex the processing was, where each roaster positions itself, and what the packaging costs, the same stack of factors covered in why specialty coffee costs more in India. The estate name is constant across that entire range, which means origin information has become abundant while a reliable price-to-quality signal has not. The estate label does little to tell a buyer whether the premium in front of them is buying better coffee or just a more elaborate story.
Ratnagiri is the clearest case, not a freak one. Across ICB's catalogue, 30 estates are each sourced by three or more roasters, and the average among them appears across 3.5 distinct roast levels and 4.4 distinct processes. The widest spans all five roast levels and ten processes. Estate-sharing is structural in Indian specialty: a relatively small set of well-known farms, like Ratnagiri, Baarbara, Harley, Attikan, and Salawara, supplies a large and fast-growing roaster base. That structure has quietly shifted where roasters compete. When everyone can buy the famous estate, sourcing access stops being the edge, and what a roaster does with the green, the processing it selects and the roast it develops, becomes the thing it actually competes on.
The shape of the spread changes from estate to estate, though, which is worth noticing on its own. Attikan Estate is sourced by 12 roasters, and its 20 coffees lean washed and classic: washed (7) and experimental (7) lead, roasts cluster from light to medium-dark, and the price spread is a narrower 2.7-fold (Rs 582 to Rs 1,563). Where Ratnagiri's roasters fan out across ferments and roast levels, Attikan's converge on cleaner washed profiles. The roaster effect is present everywhere, but each estate has its own centre of gravity, a default direction its roasters tend to share and a width around that default. That width is readable in itself. A wide spread like Ratnagiri's marks an estate the market treats as a proving ground, while a tight one like Attikan's marks an estate roasters trust to deliver a dependable, recognisable profile.
The mechanism is the three-variable stack working in practice. Start with roast development. The same dense, sweet green can be taken light, preserving fruit acids, florals, and a winy acidity, or dark, developing chocolate, caramel, and body. Roast level alone is enough to move one estate's coffee from a fruit cup to a chocolate one, which is exactly the split you see across a single region's catalogue when you sort it by roast.
Then add lot and process selection. An estate running its own processing programme offers a roaster real choices: a carbonic lot, a washed lot, a double-ferment, an SL28 anaerobic. A roaster who buys the carbonic lot and roasts it light has made a fundamentally different coffee from one who buys the washed lot and roasts it medium-dark, and that difference exists before any question of skill or roasting quality enters. Each roaster is authoring a coffee through a sequence of choices, and the estate is the raw material those choices act on.
Tip: When two bags share an estate name, the estate is the least informative thing on the label. Read the process and roast level printed next to it, and the specific lot name, like YC, SL28, Culture Naturals, or carbonic. Those tell you which of the estate's many faces you are actually getting.
The reframing is simple. An estate name is the address; the roaster is the author. For a coffee you already know to be specialty-grade, the estate is the least predictive label on the bag. Roast level, process, and the specific lot tell you far more about what is in the cup. Chasing estate names as a quality shortcut misreads where the variation actually lives.
It is worth being equally clear about what this data does not say. It does not crown a best version of any estate. Different roasters buy different lots, so they are not even roasting the same green, and ICB's rated sample per estate is small: Ratnagiri has ten rated coffees, most others two to five. That is far too thin to rank one roaster's Ratnagiri above another's, and this piece does not try to. It measures how widely the versions diverge, not which one wins.
A note on the data: Estate matching assumes that a shared estate name means the same farm, which is reliable for distinctive names like Ratnagiri and Attikan. The counts come from ICB's catalogue on 30 May 2026 and will shift as roasters add and retire coffees. Ratings are too sparse to rank roasters and are used here only as context. Prices are normalised to a 250-gram pack.
The practical upside of all this sits in the directory itself. Because ICB carries every roaster's catalogue, you can see all 23 versions of Ratnagiri in one place and compare what each roaster decided to do with it. The estate tells you where the coffee is from. Comparing roasters tells you what they each chose to make of it.
