The Blog

view all

381. Introducing the farm-facing cupping form

When I published this list of New Year’s resolutions earlier this month, there was one I forgot: create a cupping form for farmers.


The SCAA cupping form remains the long-reigning standard.  But there has been plenty of talk about alternative cupping forms in recent years, from Shawn Steiman’s call for all-out sensory revolution during the 2011 SCAA Symposium, to this article in Roast Magazine profiling roasters who have made less radical modifications to existing forms for different reasons: to account for differences in post-harvest processes, align the evaluation process more closely with their particular sensory priorities, or just simplify the whole affair.  We don’t want to burn anything down, but we do think that incorporating actionable insights into traditional cupping forms could make them more useful for growers.

Cupping forms and tasting notes are market-facing and descriptive.  Growers could use a form with notes that are farm-facing and prescriptive.

One that doesn’t just tell them a coffee is good or bad, but also tells them why.

And “so, what”–what farmers can do about it if their coffee doesn’t make the grade.

Ideally, it would also give them an idea of the quality frontier of their farms–the highest scores they can realistically hope to achieve through improvements in practices they can control.

In Colombia we are working with industry allies to create one.  We started down this road a few years ago in Central America with our friends from Cooperative Coffees as part of our CAFE Livelihoods project.  We sent them samples every year and in return they sent us cupping results using streamlined forms with three columns of feedback: one for scores, another with detailed cupping notes and a third column attributing what they tasted in the cup to specific farm-level practices, with particular focus on defects, year-on-year changes in cup profile and potential for improvement.  Here are some excerpts from the third column of a 2011 Nicaragua report, with the parts I consider particularly useful in boldface.

A loss from last year’s score.  The cups were very vegetal and phenolic.  Flavors potentially point to improper drying.


Greater insect damange and presence of sour beans produced cups that were very unclean and aged tasting. Potentially developed mold from improperly dried coffee? Very dirty compared to last year’s samples.


Nice improvement over last years sample. Grape flavors dominate and make this drink more like a nice red wine than a coffee. Great potential for this coffee.


We are picking up in Colombia where we left off in Nicaragua, trying to be more systematic this time around in targeting the recommendations cuppers make to growers.  Counter Culture Coffee, cited in the Roast article for its record of innovation in sensory evaluation, is leading the charge, along with five other U.S. importers and roasters advising our Borderlands Coffee Project.

Across the top row of the form pictured above are the traditional SCAA sensory evaluation categories and a place to register each coffee’s overall score.

The innovation starts in the second row, which identifies variables farmers can control on the farm and at the mill, with space for cuppers to scrawl comments related to each–congratulations for good practices that come through in the cup and recommendations that may help the farmer do better next time.  Finally, the form is anchored in the lower-right with a space for each coffee’s “potential quality frontier”–the top score a cupper feels a farmer is capable of if he or she were to act on all the enclosed recommendations.

This final box helps roasters identify for special attention growers who have potential for big gains in quality or whose upper bounds reach stratospheric levels.  It also helps inform the decision about how a farmer should allocate one of his or her most important resources: time.  A grower who may be producing coffees that score 80 points with an upper bound of 82 may decide the additional labor associated with a modest improvement in cup quality may not be worth it.  But a grower whose coffee is scoring 84 with a quality frontier closer to 89 may decide to act on the cupper’s recommendations and seek quality premiums.

Four caveats.

First, this is not to suggest that coffee associations don’t need cuppers perform market-facing, descriptive sensory analysis as part of their work to link smallholders to markets.  They do, of course.  But for the average smallholder grower who will not become a Q-graded cupper, an expanded cupping form that includes actionable information to improve cup quality may be more valuable that one that doesn’t.

Second, this kind of form may have the most value when it is used in the context of a direct relationship with growers–one in which regular visits to farms to see the conditions of production, harvest and processing up close complement far-off cuppings back in the lab.

Third, getting this right will take a lot more time and a lot more iterations.  It may require new kinds of training for cuppers.  And, of course, there is no guarantee of success.  The one pictured above is already the fourth iteration of the original draft form pictured below–the “Double Cupping Form” created last September by Counter Culture’s Tim Hill.  He used it for three months before scrapping it and starting over.

Fourth and perhaps most important, this whole enterprise assumes that cuppers have a deep understanding of farming and milling processes, the experience needed to attribute cup characteristics to specific farm-level practices, and ultimately the ability to generate appropriate recommendations.  It is a big assumption.  But until WCR gets its sensory program off the ground, the best thinking about what practices drive coffee quality probably lies dispersed in the minds of coffee buyers who have logged millions of muddy-boot miles in the coffeelands.

Feedback, please!


  • Thomas Oberthur says:

    Dear Michael. As usual great ideas. I would like to add another one or two caveats ….

    Caveat #5. The power of such approach will increase significantly with more data captured, the access to these data and the transparent interpretation of them. Hence, it is critical that data mgt ability has been put in place beforehand. Which in turn brings us to the question of institutional capacity and arrangements within the supply chain and within growers organization. The identification of cause (management) effect (quality) relationships becomes incrementally clearer with more information available for analyses and interpretation. Obviously the off shoot is that too little information can lead to wrong interpretation, and recommendations based on erroneous interpretation are likely to have long lasting negative impacts on growers willingness to change.

    Caveat #6. Growers ability to change management practices is obviously restricted by their ability to harvest, and process post harvest batches that have been produced using a certain set of specified conditions / practices independently from other batches. For example, east facing slopes versus west facing slopes, or beans from the lower part of the tree versus beans grown in the upper parts of the bushes. That ability to process separately is a function of availability of labour and mill capacity. Both unfortunately, is not always available as one would want it …

    Cheers. Thomas

    • Michael Sheridan says:


      Thank you for your words of encouragement and caution. Both your caveats resonate with me, but particularly the first–one can see how easily a well-intentioned exercise can have adverse unintended consequences. Also, I wonder how long the list of caveats has to get before we decide the risks outweigh the possible rewards. Is there a reason this hasn’t caught on sooner?


      • Thomas Oberthur says:


        In my opinion, implementation of such management support is overdue. It is an extension of GAPs in the field to make them more responsive to actual market demands AND potential market opportunities.

        Awareness of caveats does not constitute risks. Addressing caveats is needed to reduce implementation risks.

        Why hasn’t it happened? There may be several reasons. One that stands out on my list is the lack of clearly articulated market demand … The mechanisms, as currently implemented, that the specialty industry is using for the identification of supply are not conducive for such an approach. Evaluating cupping tables full of samples (be it at competitions, auctions, in the lab, in the sales room) is a highly effective to search for that elusive top quality cup. This mechanism does not support the systematic development of a great cup.

        To enable such development the feedback loop between the cupping table and the field has to be established. A precursor is the quantification of additional value that can be generated from such information feedback – upstream and downstream of the chain (I know that you started to do that). Then, systematic linkages between quality and management can to be established.

        These cause effect associations may be rather crude to start with, but surely, they become ever more refined with more information from the cupping tables linked to the field. Then growers and their partners can implement innovative management practices at reasonable risk (the en vogue term might be inclusive within supply chain experimentation).

        What’s the stumbling block obstructing such change? We have the information technology to support the process of linking field to cupping table to roaster (and to consumer). It seems someone needs to underwrite an initial pilot (that is your field of expertise), or we wait until supply of high quality product becomes more limiting (maybe it never will?). Until then, the cupping table is likely to remain only a mechanism to identify the outstanding cups, and may not evolve into a complementary tool that supports the systematic development of high-quality beans.

        Kind regards.


        • Michael Sheridan says:

          Thanks, Thomas, for the continued discussion and enouragement.

          Two more thoughts from my end.

          First, it strikes me that this whole exercise might be most effective in the conext of boutique exporters who separate relentlessly at origin–farmer lots, day lots, etc. Being just one step removed from the farm they are well-positioned to make reliable linkages between what they are seeing on the farm and tasting in the cup. The further an actor is downstream, the greater the informational burden on upstream actors (they need to convey what they see reliably and at a reasonable cost) and the less likely they will be able to cup on a lot-by-lot basis. Once coffees from multiple farms and multiple mill days are blended for sampling and shipment, it becomes all but impossible to track defects in the cup back to the farm level and deliver the kinds of actionable feedback the farm-facing form proposes to generate.

          Second, I was thinking of you when I wrote the words “quality frontier” given the amazing work you did here in Colombia with CIAT literally mapping coffee quality frontiers in Cauca and Nariño–identifying agroecological niches with a high probability of producing high-quality coffees. I wrote in the original post that a farm-facing form like the ones we are working on in Colombia is most valuable when a cupper has an intimate knowledge of what is happening on the farm and at the mill. Would it be safe to say that it would even be more valuable when accompanied by this kind of research? How would you see a cupper of coffee from Nariño incorporating your work mapping quality niches with the estimate of quality frontier?


  • Michael:
    Thanks for this thoughtful piece on cupping forms. I’ve appreciated some of the alternative forms Willem Boot has created. I suspect David Griswold must have some variations in his field-ready Ipod system, too.

    I like how the form you’re proposing tries to improve the connection between cupping scores and farming/processing practices. This issue points to a deficiency I noticed in the SCAA’s official green grading book. When I first got that book and took the class, I thought there would be much more there linking the various defects to farm practices. It’s pretty vague.

    You seem to imagine buyers being the primary users of this form, and hint that a few individuals at cooperatives could use it. Maybe I just misunderstood. It seems like the power in the concept of this form is to target use of such a form by producers and washing station/mill owners. Coffee buyers have logged millions of muddy-boot miles, but I suspect the growers have logged more. Somehow perhaps a system could generate their interest and thereby enhance their participation in measuring and tracking what is happening in the field and at the mill (the data gathering that Thomas O. talks about).

    You say you are working with allies in Colombia to develop the form. I’m guessing some of those allies are cooperative owners and smallholder farmers? What are they saying about it?

    Great highlight about producers’ allocation of time.

    Best regards,
    Ruth Ann

    • Michael Sheridan says:

      Ruth Ann:

      Thank you for your comment. I was not aware that Willem had devised alternative cupping forms–would love to see them if you can send them along, especially if they include fields for comment on farm-level practices.

      In terms of the “who?” of the forms, you are right–if it works (a big IF), and if buyers and sellers can get calibrated in their farm-facing approach just as they do in their market-facing approach (an even bigger IF), then it should be used wherever coffee is cupped. The closer that cupping happens to the farm, the better, as the whole idea is to create a more robust feedback loop. As you know, many cuppers at origin are growers themselves, and all are much closer to production and milling processes than U.S.-based importers and roasters. In that regard, they may be better positioned to turn farm-level insights into quality improvements than their downstream trading partners.

      The participants in our Colombia effort are all U.S.-based buyers at this point, but over the next year we will be building and staffing two washing stations where cuppers will perform lot-by-lot evaluations–an arrangement that will position us well to move this process closer to the farm. Without question, for this to work, growers and other origin-based actors would need to send a pretty broad band of data downstream with their coffee.


  • This sounds really amazing. Slate would love to participate if the opportunity ever arose!

    • Michael Sheridan says:


      Thanks for your interest–let’s talk!

      Not sure where a collaboration would begin, to be honest. You could do as Tim Hill has done at Counter Culture–take one of the forms here or create your own, dive in and share your thoughts on the process as it evolves–what’s working, what isn’t, what can be improved.

      I think to do this well we would need to create a pretty systematic process where we harvest and sample coffees that have a range of defects (underripe, overripe, rust-affected, CBB-affected, ferment, etc.) and represent different points in the continuum in terms of drying time, shade cover, fertilizaton, etc., cupping them together with a panel of roasters committed to this process to begin to develop associations of specific cup profiles with specific practices on the farm and in the mill. A heavy lift, as they say!


  • Hello Mr. Sheridan, I’m a coffee agronomist from Brazil and a coffee amateur grader too. I work in a medium farm in a traditional small coffee producer city in São Paulo state, I believe this form is easier to understand and use helping me to score my coffee for different areas and varieties we have in the farm.


  • Great post Michael and wonderful contributions by all.

    I mainly want to second the great notions put for by Ruth Ann that the ones MOST qualified to use such a form are the farmers themselves. They are the ones directly tied to processing techniques that can make or break a coffee on the cupping table. They might need some initial education about how processing techniques can affect the cup and longevity of a coffee but after that, they can take the reigns assuming they have the ability and infrastructure to taste and cup.

    One limitation that is extremely important to mention, however, is that farmers don’t always have the ability to taste the coffee throughout the year, as we roasters do. How well a coffee is processed is directly related to how stable the coffee will be after processing and this is where great processing will really dominate, 4-6 months after a coffee is harvested.

    I also second the notion put forth by others that there are really very few cuppers who are educated enough with processing conditions to make good, solid and safe recommendations. Every farmer or cooperative is set up differently and every year is going to be different in terms of climate as well. There is so much information needed so a cupper can make educated deductions based on what they are tasting and based on what they know and don’t know.

    I hear a lot of talk about improving quality but I find that the reality for most organizations is that they have a tough time getting remunerated for the work it takes to improve quality. There are certain financial brackets (differentials) for every country and it is really difficult for producers to bust out of those realities, no matter how good your coffee is. Before changes are made, it is important to know what the costs will be, both to the farmers in terms of their time and risk, as well as if it is worth it to make such changes. If a farmer needs to expend 25% more effort to increase quality 2 points, they need to find a buyer willing to pay that. For a two dollar FOB coffee, that’s .50 cents and most buyers I know can’t justify it.

    Confusing this matter more is scoring. I am of the opinion that the SCAA scale does indeed need to be looked at and revised. I am always surprised to see how many of the best coffees in the world will score 86-87 on the SCAA scale. Effectively, this scale stops at 90. Why create a 100 point scale that stops at 90? Because of this, there is a lot of compression in scoring in the 83-84 range where I feel like that we would do better to broaden the scale to a full 100 points. This would flesh out scoring a lot and better reflect the reality of scoring which is to say that a 90 point coffee is not an A-, it’s an A+.

    The notion that every country should be scored with different relative standards also makes no sense to me. A scoring scale will make way more sense if it has universal standards that apply to all coffees, no matter where they are from. A 90 point Brasil should be theoretically as good as a 90 point Ethiopian or Kenyan. Why would each have its own standards? It is very confusing to farmers, buyers and consumers to apply different scores to different coffees based on where they come from. Theoretically, based on SCAA standards, a cupper can score the same coffee in two different ways depending on if he thinks it is a Guatemala or a Brasil. That is super confusing. Kudos to Counter Culture for breaking out of this model and paving the way for scoring coffee differently, on a universal, 100 point scale.

    That’s all for now. Thank you Michael for engaging these discussions. You are a very important resource for all of us in this industry. We still have so much to learn.


    • Michael Sheridan says:


      Thanks for the generous comment and completing the 1-2 punch from Kickapoo. The message you and TJ are conveying about how ill-suited roasters are to lead this kind of a process comes through loud and clear. As does the caveat about tasting coffees over the course of the year to see how they hold up in the cup over time. My ongoing coffee education is schooling me to link improved (slower!) drying practices with longer staying power in the warehouse, but that is something to be tested through the kind of ongoing cupping you mention. No reason we can’t do that at origin, too. Finally, and most importantly, thank you for bringing the conversation back to price. I noted in my post last week the disconnect between specialty’s message to growers (we prefer traditional varieties to hybrids) and its willingness to compensate farmers for the added risk associated with those cultivars. Unless specialty can demonstrate its willingness to pay higher prices for what it regards as the good stuff, it is paving a path to a future in which quality-differentiated coffee is the domain of highly capitalized estate and craft producers.


      PS: I am compelled by your arguments for reengineering the scoring system, but that conversation is way beyond my pay grade. I will leave that to the Kickapoos and Counter Cultures of the world. M

  • TJ Semanchin says:


    I feel that we as cuppers can certainly taste it when a coffee is harvested and processed perfectly or imperfectly, but to rely on feedback from that sensory anylsis for accurate recommendations to famers in the field is too difficult and imperfect to implement.

    Rather, what I feel is the best path forward and is already emerging as best practice in the field, is to work with and support internal capacity within cooperatives that are able to assess and evaluate quality throughout the post harvest handling chain. So what I propose is to de-emphasize cupping reports (still a crucial analysis tool) and promote quality evaluation starting at the point of delivery of parchment from farmers to the cooperative’s bodega. Inspection of the quality and humidity of the parchment is already and certianly should be the norm for a co-op, what is often lacking is the recording of data from this analysis, and the subsequent feedback to the producer after the harvest. Included with this initial data should expected yield of parchment to export quality green coffee, the results of the visual green evaluation (e.g., number of defect), the roasted evaluation (e.g., number of quakers), and finally the sensory evaluation of the cupping report with emphasis on faults resulting from processing errors. Recomendations to individual farmers can then be offered based on all this data. Also, from this comprehensive evaluation the co-op can track progress with a specific farmer from year to year and perhaps reward that progress with quality incentives (with the caveat this is culturally inappropriate for certain farmer groups). Another benefit with this approach is that with best quality practice recommendations to individual farmers, there is a mechanism in place that could also offer recommendations for increased yields and other cost saving, or profit increasing, methods.

    This requires substantial internal capacity within cooperatives, but it could be a goal as buyers and responisble trading partners to promote this internal infrastructure with trainings, equipment, or other resources.

    Thanks for digging so deep into the on-the-ground specifics of how to best serve farmers on their path to higher incomes.

    • Michael Sheridan says:


      Thanks as always for the thoughtul and insightful comment. I love just about everything you have to say, especially the idea of the emerging best practice of building these skills as close as possible to the farm level. In our work in Nariño, where growers have proven pretty allergic to cooperatives (the only FT coop in Nariño claims just 280 of the department’s coffee growers, who number more than 38,000 smallholders) we are building two washing stations. We see them as ideal platforms for precisely the kinds of detailed mill-to-farm feedback you are talking about. So many coop mills are dry mills only, so there is less direct observation of harvest point by the team working at the mill. By incorporating the washing process in our model, we add another layer of quality-focused oversight when cherry is received. From there, everything you say resonates loudly–generating feedback to the farm based on both cherry selection and graded green coffee, maintaining separation of farm and day-lots to close the feedback loop. All of this is an easy sell…The question that remains in my mind is how to get the roaster engaged constructively. This may be one clear case where quality norms are co-constructed. Or it may be that in this case, quality training goes with the tide–instead of market-based graders and roasters training farmers and coop staff upstream, folks closer to the ground can share what their quality-focused work reveals about the links between specific farm practices and cup quality.


  • There is a need for more professional cuppers in coffee countries. Cuppers provide information that help roasters choose green coffee. However, cupping forms focus on characteristics in the coffee cup not on the way the beans were grown, processed, stored or roasted. The fact that some cup characteristics are indicative of poor bean management has led to a fallacy that cuppers can tell farmers how to improve their quality. Cupping is not a farm management tool.

    Typically, professional cuppers have little practical knowledge about farming and processing techniques and, most importantly, they cannot connect their cupping assessment to the farmer’s practices. A two week trip to origin isn’t the same as a life time of planting, pruning , fertilizing, growing, picking and processing coffee.

    A written cupping assessment noting low aroma, high acidity, or medium body isn’t going to change farmer behavior. A Q cupper can tell a farmer that her coffee was fermented but he cannot tell her what caused the ferment. This is the same concept stated above by TJ Semanchin, “I feel that we as cuppers can certainly taste it when a coffee is harvested and processed perfectly or imperfectly, but to rely on feedback from that sensory analysis for accurate recommendations to farmers in the field is too difficult and imperfect to implement.”

    Cupping is an assessment of coffee after it has been picked, pulped, dried, rested, hulled, sorted, ground and brewed. The distance between coffee beans on a tree and a cup of coffee is almost infinite. Identifying a defect in a brewed coffee cup doesn’t tell anyone where the defect occurred or how it can be corrected.

    For many farmers cup quality is a remote concept. Some farmers who sell cherry or parchment coffee have never tasted their own coffee. It is a misconception that farmers need to know how their own coffee tastes. If a farmer only sells cherry coffee, she doesn’t need an assessment of her brewed coffee – she needs an assessment of her coffee cherry. Teaching a cherry farmer how to cup coffee is as useless as teaching a homeless person how to cook duck a l’orange. Farmers need to know what the characteristics of their coffee beans are at the stage that they sell them and, most importantly, how they can become more profitable

    The same analysis is true for the parchment farmer and the green farmer. As stated by TJ Semanchin, ” I propose to de-emphasize cupping reports (still a crucial analysis tool) and promote quality evaluation starting at the point of delivery of parchment from farmers to the cooperative’s bodega.”

    Most coffee farmers are not driven by fame and glory. A high cupping score is meaningless when a farmer is fighting bugs, fungus, high fertilizer cost, organic and inorganic chemical cost, weather, thieves, terrorists and wild animals. Farmers are in the coffee business to survive and make a living. Telling a farmer he can get a high cupping score when his children are hungry is a meaningless communication. Telling a farmer he can make more money with better quality coffee is not. Unless farmers receive specific information about their good and bad practices which they can monetize, they are not going to make changes. Instead of a cupping score farmers need a quality assessment that reflects both the good and bad characteristics of their coffee at the stage that they sell it, how to improve those characteristics and the monetary benefit that will result.

    It shouldn’t be difficult to design assessment forms for cherry, parchment, green and roasted coffee beans. Cupping is an assessment form for brewed coffee beans which works well for roasters and consumers. Farmers need assessment forms which relate to the product they sell not to another product created down the production line – miles away from the farm. Most importantly, they need a simple cost benefit analysis of making improvements in their coffee quality.

1 Trackback

Leave a Reply

Your email address will not be published.

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

Subscribe to this comment feed via RSS