How India's Coffee Processing Mix Is Changing
At commodity volume, Indian arabica has been wet-processed for over a century. The specialty roaster catalog shows something different: experimental methods now represent more than 20% of labeled coffees in a directory of 1,240 entries, up from roughly 7–8% in an earlier snapshot. This is a field guide to the two-track processing shift: what the data shows, where it is happening, and what drives the divergence between the commodity baseline and the specialty catalog.
Indian arabica has long been understood as a washed origin. The Coffee Board of India built that identity into its grading system: "Plantation A/B" means wet-processed arabica; "Parchment" means wet-processed robusta. USDA data confirms the production reality: roughly 80% of India's arabica and robusta output is wet-processed, a share stable for decades.
ICB's specialty catalog is less settled. Of 1,240 Indian coffees listed from active D2C roasters, 832 carry a clear process label. Among those, anaerobic fermentation and related methods — anaerobic, carbonic maceration, double fermented, and lots tagged simply as "experimental" — together account for 253 coffees. That is 30.4% of labeled lots and 20.4% of the full directory. Washed sits at 305 coffees, or 36.7% of labeled lots. The specialty catalog and the commodity statistic describe different versions of the same production base.
Is India actually changing its processing approach, or is the catalog a marketing layer over an unchanged production base? The answer involves two separate tracks operating in parallel. Understanding which track a given statistic comes from determines which conclusion holds.
The commodity baseline: why India is still 80% washed
India's processing baseline was shaped by infrastructure. The Coffee Board's grading nomenclature encodes this logic: "Plantation A" is washed arabica at screen 17 or above; "Plantation B" covers screen 15–16; "Cherry AB/PB/C" is natural-processed arabica. These weren't just grade names. They were delivery specifications for commodity export markets, and the wet-mill infrastructure built around them has been in place for generations.
The Western Ghats made washed the natural default. Water access across Chikmagalur, Kodagu, and Wayanad is reliable from monsoon rainfall and river systems. Existing concrete fermentation tanks, pulping machines, and drying patios represent capital that most estates are not replacing. At commodity scale, washed processing also produces more consistent parchment moisture and fewer defects per lot.
Robusta followed the same path. The Coffee Board launched Robusta Kaapi Royale (RKR) in the late 1980s, making fermentation a required step for its first dedicated fine robusta designation. Even the species historically associated with dry processing moved toward wet-processing once a premium grade existed.
On the Coffee Board grade system: Two categories define India's processing baseline. Parchment coffee — wet-processed arabica or robusta — accounts for roughly 80% of arabica output. Cherry coffee, the dry-processed equivalent of natural, accounts for the remaining 20%. The specialty catalog's diversification is measured against these base rates.
One region departs from this pattern. In Araku Valley and the broader tribal cooperative belt of Andhra Pradesh and Odisha, natural processing is more common than the Western Ghats average. The Eastern Ghats post-harvest climate is drier. Open-air cherry drying is more feasible there without the humidity and mould risk that complicates outdoor drying in Karnataka and Kerala. Araku's processing profile reflects geography as much as specialty positioning.
What the specialty catalog actually shows
ICB's directory of 1,240 coffees from active D2C roasters is not a representative sample of India's production. It is a catalog of what specialty roasters have chosen to sell: single-origin, traceable, differentiated lots rather than commodity blends and standard estate grades. That context matters for reading its processing numbers.
Among the 832 coffees with clear process labels:
Washed is still the largest single category. But the experimental cluster is close, and the gap has narrowed since earlier snapshots. The processing comparison article published on ICB worked from a catalog of 878 coffees. In that data, washed held 21.3%, natural 10.7%, and anaerobic roughly 7.6%. The current catalog is 41% larger. Experimental methods have grown faster than the catalog itself. Among clearly labeled entries, washed's share (36.7%) and the experimental cluster's (30.4%) are now close enough that process diversity, not washed predominance, defines the specialty catalog.
The 408 unlabeled entries are part of the picture too. Many represent genuinely hybrid or multi-stage processes — a coffee fermented anaerobically and then dried as a natural doesn't resolve cleanly to either label. Some roasters use "experimental" as a catch-all for yeast inoculation, co-fermentation, and extended fermentation in a single tag. And some catalog entries are commodity-adjacent products (ground coffee, blends, roast-to-order bags) that were never going to carry specialty process metadata. The unlabeled 32.9% is not simply missing data.
Why the mix is shifting: the pricing premium
The direct driver of specialty catalog diversification is price. Experimental processing commands a documented premium over washed lots, and that premium is large enough to cover small-batch fermentation work.
ICB's catalog data shows median prices per 250g by process:
| Process | Median price (per 250g) |
|---|---|
| Washed | ₹615 |
| Honey | ₹730 |
| Natural | ₹749 |
| Experimental | ₹890 |
| Anaerobic | ₹900 |
| Carbonic maceration | ₹900 |
Anaerobic and carbonic maceration lots sit 46% above the washed median. For estates producing small batches, sometimes under 100 kg of green coffee, that per-kilo difference covers sealed fermentation tanks, temperature monitoring, extended fermentation windows, and the additional labour of careful micro-lot management. Lots at ₹1,200–1,800 per 250g, the highest-priced tier, have continued to find buyers. That the top tier keeps selling suggests buyers are treating fermentation protocol — not just origin — as worth paying for separately.
The demand came from roasters before it came from estates. Specialty cafes and roasters in Bengaluru, Mumbai, Delhi, and Kolkata moved from requesting clean, balanced profiles to asking for fruit-forward and fermentation-forward cups. One second-generation estate operator noted that five years ago, buyers wanted standard varieties; now they request coffees that taste of apples, jackfruit, or bananas. Estates followed that demand. Experimental processing developed as a response, not a supply-side innovation searching for a market.
On scale: Experimental lots are typically 2–5% of an estate's total output. The model is additive: estates still process 95–98% of their cherries through the standard wet mill. No significant Indian estate has converted primary production to anaerobic or carbonic maceration at volume.
Where the shift is happening: a regional map
Experimental processing adoption is uneven. ICB catalog data shows a clear geographic pattern, with specific sub-districts leading by experimental share and by absolute count.
Sakleshpur leads by share at 54% of its catalog. Carbonic maceration lots concentrate here — six of the region's entries are labeled CM specifically, the densest CM concentration in any sub-region. Baba Budangiri holds the highest absolute count (35) relative to catalog size (27%). Ratnagiri Estate, which adopted custom-built stainless-steel fermentation tanks modeled on Colombian anaerobic practice, is the primary driver. It is the most process-diverse estate in the ICB directory, with washed, natural, honey, anaerobic, carbonic maceration, and double-fermented lots all cataloged.
Kodagu/Coorg shows the highest carbonic maceration share of any region (10 CM coffees of 97 total), which suggests CM specifically, rather than anaerobic generally, has become Coorg's characteristic experimental method. Chikmagalur has the largest raw count of experimental coffees (62) because its overall catalog is the largest, though its experimental share at 21% is proportionally lower than Sakleshpur or Baba Budangiri.
These differences aren't arbitrary. Sub-regions with early-adopter estates that invested in fermentation infrastructure and built direct supply relationships with specialty roasters have higher experimental shares than geographically adjacent areas without that estate-level investment. Araku Valley tells a different story. Tribal cooperative production prioritises volume consistency and certification. The economics of experimental micro-lots don't transfer easily to the smallholder cooperative model, and the region leans toward naturals and washed lots rather than anaerobic fermentation.
How the additive model works
An estate processing 200 tonnes of cherry through the wet mill during harvest might set aside a single picking day — two to five tonnes from a specific plot — for sealed fermentation. That batch fills one or two stainless-steel tanks, ferments for 36 to 72 hours under controlled temperature, gets hulled separately from the main parchment batch, and dries on raised beds while the rest of the harvest moves through normal channels. The resulting green coffee might be 100 to 300 kg: too small to register in export statistics, but enough for two or three specialty roasters to offer as a named micro-lot.
The estate's total wet-processed output barely changes. But it now appears in the specialty catalog under a different process label. The catalog and the production statistic are both accurate. They are just measuring the same farm differently.
The risk-sharing model that made this viable came from roasters committing to purchase entire experimental lots regardless of outcome, absorbing the commercial risk of a failed ferment in exchange for exclusivity or first access. Blue Tokai's Producer Series, launched in 2020 and now releasing rotating nanolots annually, formalised this structure. Black Baza Coffee produced over a dozen microlots in 2025, fermented with kombucha cultures and wild ginger foraged from the same areas as their smallholder sources. All of those lots sold at premium pricing. Consumer demand proved broad enough to sustain the model beyond the early-adopter roasters.
India's fermentation inputs: local and specific
India's shade-grown coffee grows alongside pepper, cardamom, arecanut, and tropical fruit trees. That polyculture generates a microbial environment that differs from monoculture coffee farms in Brazil or East Africa, and it shapes what experimental fermentation looks like in practice.
When Indian estates co-ferment — adding biological material to fermentation tanks to guide microbial activity — many draw on this local ecosystem. Black Baza's use of wild ginger foraged near their smallholder sources and on-farm kombucha cultures is hyper-local in a way that can't be replicated off that specific land. Co-fermentation in some South American origins often relies on imported passion fruit pulp, mango, or commercially produced yeast cultures. The ambient microbial load of a polyculture estate in Coorg or Chikmagalur is a different starting point, and it produces different fermentation outcomes.
Yeast inoculation using specific lactobacillus or Saccharomyces strains is also present, particularly at estates with more formal microbiology investment. The distinction between guided fermentation (inoculated, temperature-controlled, documented) and ambient fermentation (natural microbial load, less controlled) matters for reproducibility. Inoculated lots tend to be more consistent across harvests; ambient fermentation produces more harvest-to-harvest variation. For a buyer assessing an experimental Indian lot, that distinction matters as much as the process label itself — two lots labeled "anaerobic" on the same estate can diverge significantly if one used documented temperature control and the other did not.
The labeling problem
Of 1,240 coffees in the ICB catalog, 408 carry "other" or "unknown" as a process label, and 55 are tagged "washed_natural" without specifying the split. That 28.5%+ is not missing metadata — it reflects a market diversifying faster than labeling practice can follow.
Some entries represent hybrid or multi-stage processes. A coffee fermented anaerobically and then dried as a natural doesn't belong cleanly to either category. Some roasters use "experimental" as a marketing umbrella for yeast inoculation, co-fermentation, and staged fermentation together, without specifying which. And some entries are commodity-adjacent products where process metadata was never the point.
When reading Indian specialty bags: "Experimental process" more often indicates controlled fermentation with a documented protocol than improvised or informal work. Look for fermentation specifics on the roaster's website: hours, temperature, inputs. Lots where the roaster can describe what happened in the tank are generally more predictable in the cup than lots where "experimental" is the only available information.
Reading the catalog: process as a buying signal
Washed lots (305 in the catalog) are the largest single category and the most consistent entry point for tasting a specific estate's terroir and a roaster's development decisions. Natural and pulped natural lots (159) have grown proportionally since earlier snapshots, with sub-region mattering for flavour expression: Western Ghats naturals dry under different humidity and temperature conditions than Araku Valley naturals and produce different fruit intensity levels. Honey lots (81) occupy the middle ground, adopted mainly from the mid-2010s onward, typically producing cups with more sweetness and body than washed but less fermentation intensity than anaerobic lots.
The anaerobic, carbonic maceration, double fermented, and experimental cluster (253 combined) is the fastest-growing premium tier. Flavour in this category is more protocol-dependent than terroir-dependent. Two anaerobic lots from the same estate can differ significantly based on fermentation duration, temperature, and inputs. These coffees also tend to be more sensitive to extraction: compounds produced during extended fermentation are generally more soluble, so small brew-variable changes produce larger cup differences than equivalent adjustments on a washed lot.
The pricing tiers — ₹615 median for washed to ₹900 for anaerobic — are a reliable proxy for processing complexity before the bag is opened. That premium holds consistently enough across roasters and regions to suggest it reflects real input cost differences rather than positioning alone.
