The Blog

view all

438. The CRS Colombian Varietal Cuppings series adjourns (for now)

Over the past six weeks, this blog has been devoted exclusively to the CRS Colombian Varietal Cuppings–a series of comparative cuppings of Castillo and Caturra samples from our Borderlands project involving leading roasters and importers in the United States, Europe and Australia.  Even when I took a week off from reporting on the results of the cupping events, I dedicated the blog to a discussion of their implications for agrobiodiversity, smallholder livelihoods, policy and industry buying practices.

For readers into this series, the blog’s focus over the past month and a half has been a source of excitement.  For readers who don’t share their enthusiasm for the issues at stake here, well, not so much.  There is welcome news today for readers who fall into both camps.

Tomorrow the events continue with a cupping in Sydney, but the CRS Colombian Varietal Cuppings series will adjourn today.  We will resume coverage in April, when the jury in the Colombia Sensory Trial renders its verdict and brings all-new-and-improved perspectives to the conversation that is just getting started here.


At least two dozen roasters and importers on three continents (and one coffee publication) participated in the CRS Colombian Varietal Cuppings.  I want to thank them here for giving so generously of their time and talents to this process:

  • Atlas Coffee (Drew Billups, Chris Davidson, Al Liu)
  • Campos Coffee (Benjamin D’Emden)
  • Coffee Supreme (April Fahey)
  • Conscious Coffees (Mark Glenn)
  • Counter Culture (Katie Carguilo, Tim Hill, Kim Elena Ionescu, Hannah Popish)
  • Crop-to-Cup (Ben Heins)
  • Five Senses Coffee (Jacob Ibarra)
  • George Howell Coffee (George Howell, Jenny Howell)
  • Intelligentsia (Alexandria Barnes, Alex Burbo, Jay Cunningham, Chris Kornmann, Amanda Seaver, Geoff Watts)
  • Ipsento (Tim Taylor)
  • Kickapoo Coffee (Caleb Nicholes)
  • Madcap (Ryan Knapp)
  • Marvell Street Coffee (Nicolas Kypreos)
  • Mecca Espresso (Mark Howard, Tuli Keidar)
  • Paramount Coffee (Dylan Johnson)
  • Proud Mary Coffee (Lucy Ward)
  • Reuben Hills (Nick, Russell)
  • Seven Seeds (Aaron Wood)
  • Square Mile (James Hoffmann, Anette Moldvaer)
  • Stumptown (Steve Kirbach, Adam McClellan, Craig Olander)
  • Swiss Water Decaf (David Kastle)
  • The Coffee Review (Kenneth Davids, Jason Sarley)
  • Tim Wendelboe
  • Tony’s Coffee (Andrew Bowman, David Yake)
  • Virmax (Alejandro Cadena, Pedro Cotamo, Luis Gutiérrez, Fred Lullfitz)


The participants identified here cupped between five and 22 sample pairs.  Each pair consisted of one Castillo sample and one Caturra sample from the same farm, where they were grown, harvested and processed under similar conditions.  Our careful selection process on the farm reduced the degree to which differences in cup quality could be attributable to differences in agroecology, husbandry, harvesting and processing and increased the degree to which they could be attributable to genetics.  After we collected the samples, we cupped them twice and asked our friends at Virmax to cup them again before approving them for shipment.  This screen was used to eliminate samples with defects or attributes–principally pronounced frutiness–that was the result of post-harvest processing and not an expression of the coffee’s genetics.  Samples that were approved–22 sample pairs, or 44 samples in all–were shipped bearing only randomized codes.  We didn’t reveal the identities of the samples to participants until after they cupped them so that their results couldn’t be influenced by any biases they brought to the cupping table.  I surveyed participants electronically before and after their cuppings in an effort to identify preconceptions about the varieties and possible changes in attitudes or buying policies as a result of the cupping.  For each participant profiled on the blog to date, I have also conducted a phone interview.


The results–and implications–of the CRS Colombian Varietal Cuppings that stand out to me so far are these:

  • Split decision.
    Tim Hill called it a decisive victory for Caturra.  Geoff Watts gave the nod to Castillo, and not by a narrow margin.  But other judges in this bout turned in scorecards that suggest a split decision. George Howell and Tim Wendelboe called it a draw.  George said he couldn’t tell one variety from the other, and Tim W. called it right down the middle: “Sometimes Castillo will taste better than Caturra from the same farm and sometimes the opposite.”  The head-to-head contest, in other words, has not delivered the kind of simple clarities that a knockout or a unanimous decision might have.
  • Castillo shed some unwelcome baggage and earned another chance with skeptics.
    George Howell expected to taste “the tail of the devil” in his Castillo samples.  He didn’t.  What’s more, he awarded two 90-point scores to Castillo samples and applied to them the kinds of descriptors–sparkling acidity and elegant flavor–not often reserved for Castillo.  Geoff Watts is one of the most experienced buyers in specialty, and he really responded to Castillo’s bright acidity.  Stumptown’s Adam McClellan still prefers Caturra, but conceded that Castillo’s acidity was more multidimensional than he had thought it could be, with malic and tartaric acidity showing up where he expected only citric.  Even if skeptics of Castillo aren’t ready to start buying it, they are ready to give it a second look.  “It opens me to tasting more Castillo samples,” George Howell concluded.
  • The farm: The devil is in the details.
    This process was set up to generate a rather narrow finding: whether one variety presents a clear advantage over another on the cupping table.  If there is a devil in the process, it may not be in the aftertaste of Castillo but in the details that were not part of this process.  If a grower is going to understand which variety delivers the bigger return on investment, more data are needed from the farm, where the differences between the two varieties in terms of cost of production, yields and risk must be quantified.
  • The market: Incentives.
    But the need for greater quantification isn’t limited to the farm.  The difference in cup quality must also be quantified.  If one variety scores higher than another, how much is that difference worth in the marketplace?  How do points on the cupping table translate to dollars and cents on a contract?  It may be necessary for buyers who prefer Caturra to rethink the incentives they are offering to farmers to grow it.
  • Policymaking: What is the appropriate use of public resources?
    Finally, Colombia’s coffee institutions have aligned squarely behind Castillo as the country continues its recovery from a coffee leaf rust epidemic.    In order to access certified seed and subsidized credit, growers must renovate with Castillo.  Private exporters whose customers prefer Caturra have begun to provide seed and agronomic assistance in support of Caturra and other high-quality cultivars for which there is a quantifiable demand and a demonstrated willingness to pay market premiums.  The early results of those private initiatives are encouraging.  There is a clear logic behind limiting the use of public resources to the highest-yielding, most-resistant, lowest-risk varieties.  There may also be a case to be made for public support for other varieties on a smaller scale for growers who seek higher rewards and understand the risks.  If that conversation happens, I, for one, hope it wil be nested in the broader emerging discussion on the appropriate role of the state in the coffee sector.


The CRS Colombian Varietal Cuppings are related to but distinct from the Colombia Sensory Trial.  Both rely on the same samples collected from 22 farms participating in our Borderlands Coffee Project in Nariño, Colombia, and both are based on blind cupping protocols.  But the Colombia Sensory Trial is more rigorous.  It is based on centralized panels where variability in factors that affect cup quality can be controlled, including roast levels, grind, water quality and temperature, scoring systems, storage, etc.  It will involve exhaustive analysis of results by economists, sensory researchers and practitioners.  And it will produce the kind of report that can stand up to academic and scientific scrutiny.

So, if we knew the Colombia Sensory Trial would produce rigorous results, why did we bother with a parallel (and less controlled) process at all?

  • We wanted to create opportunities for more people to participate. 
    We think that the companies and cuppers we were able to engage as panelists for the Colombia Sensory Trial are among the brightest lights in specialty.  But there was only room for eight on the panel.  This process allowed us to reach out to more buyers in more markets and get them involved directly.
  • Buyers will act based on what they taste, not what they read. 
    Regardless of what verdict the Colombia Sensory Trial jury may render in the case of Castillo v Caturra, we believe people who buy coffee are more likely to act on the basis of their own sensory experience than vicarious sensory experience.  Their attitudes about varieties and buying practices will not be changed by what they read, but by what they taste.
  • We were impatient.
    We simply couldn’t wait until 2015 to begin the conversation.  We wanted to hear from tastemakers right away and to engage others in the discussion of the results and their implications for growers, industry and policy.  We are delighted that the posts here have helped start the ball rolling.

The Colombia Sensory Trial results, which we will present during The SCAA Event in Seattle next April, will include important enhancements, including:

  • A sensory “verdict.” 
    By April, we will know what the panels told us about their preferences between Castillo and Caturra.  Whether there is even a statstically significant difference in the scores or a qualitative difference in the descriptors assigned to each cultivar, and if so, which is “better” and by how much?
  • Econometric modeling of livelihood scenarios.
    I have framed the livelihood discussion here by introducing the variables that I believe affect decision-making at the farm level: cost of production, productivity, cup quality, price premiums, total income.  The fifth is a function of the four that precede it, and the two varieties “on trial” differ in their average values across the first four variables.  But I just introduced the variables, I didn’t quantify them.  We will work with the economists at CIAT to assign average values from our field work on Borderlands to those variables so we can present some scenarios grounded in quantitative rigor during The SCAA Event.
  • Diverse perspectives.
    I continue to refer to the Colombia Sensory Trial in the sexiest way imaginable: as a decision-support process.  Our motivation in conceiving and catalyzing this process has always been to introduce more and better information into decisions currently made under conditions of considerablby significant uncertainty.  We see the results informing the varietal selection decision on the farm (which cultivar to plant?), in the policymaking process (which cultivar(s) to support with public resources and how?) and in the industry (which varieties to buy and under what conditions?).  Each of these three perspectives will be reflected in the presentation of the results: World Coffee Research Executive Director Tim Schilling and CIAT Senior Researcher Mark Lundy will present the evidence for farm-level decision-making, Geoff Watts of Intelligentsia will share an industry perspective on the process, and Luis Fernando Samper will present the view of the Federación Nacional de Cafeteros, whose policies and practices have made Castillo the dominant variety in Colombia’s coffeelands.

– – – – –

The Colombia Sensory Trial is supported by a grant from the Howard G. Buffett Foundation.

1 Comment

  • very interesting studies are being conducted to compare the two varieties grown in Colombia. As exporters of high quality coffee in Colombia whit presence in different regions of the country as Huila, Nariño, santander and cundinamarca we would be very interested to know the final results of the study to give recomendations technically supported to our producers.

    Please tell our support if anything we can help them.

    ricardo granados
    trade and Quality assurence director

Leave a Reply

Your email address will not be published. Required fields are marked *

Note: You can use basic XHTML in your comments. Your email address will never be published.

Subscribe to this comment feed via RSS