Editor’s Note: We’ve made a PDF available to you, as well as uploading it as a publicly available Cropster form, so you can choose itit’s called “Deviation Production Cupping (Royal).”

Cupping: Time Honored and Honed for Purpose

Cupping coffee is one, if not, the best-known methods for evaluating the quality of roasted coffee, and still one of the most considered methods when judging the sensorial attributes of this agricultural product.

The practice asks you to employ your gustatory receptors and, almost simultaneously, utilize the analytical tools of your cerebrum. With understanding, training, and practice, as your experience grows, your ability to report on what you taste becomes more honed, consistent and edifying to yourself and your colleagues.

And whilst the simple act of cupping has changed very little, the specialty coffee industry, driven by the global Specialty Coffee Association, has implemented global protocols, those of which are revisited and refined regularly. An excellent directive of how to set up a cupping can be found in this article, written by my colleague Chris Kornman.

Although cupping was, in the past, primarily used for defect detection, we, as members of the specialty coffee community came to rely on this practice as the method of sensorial analysis and descriptor collection.

Indoctrinated into the CQI system early, I was very fortunate to be taught by industry thought leaders from an early point in my career. I was aware, hyper-aware even, of how to cup carefully, thoughtfully and methodically (even properly?)  using the SCA Cupping Protocols, the main tool of which is the SCA cupping form.

However, as much as I admire, respect and appreciate this system, the 100-point arabica coffee grading scale, and the corresponding cupping form that enforces it, were never supposed to use them to assess more than the inherent characteristics of the coffee itself. And that’s great for assessing coffee quality. It’s an excellent tool for communicating the results of a sample roast and everything a coffee has to offer over a range of temperatures. It can communicate clearly about any defects or even just uniformity issues that may be a result of the variety, the harvesting or the processing of the coffee. What this method of cupping and form cannot do is offer more than very limited help in understanding how to assess the roast of a coffee and, more importantly, how to record data pertaining to roast consistency.

It behooves me to note at this point the fact that believing a coffee’s taste can be divorced from the way it is roasted is not just fantastical, it is fallacious. Green coffee is merely potential, it is time and temperature that makes green coffee brown and turn something distinctly unpalatable into a prized commodity and precious morning beverage. Time and temperature are the same variables used in all coffee roasting, be it sample or production.  A coffee’s flavors develop because of the roast process. The recipe, or roast profile, dramatically impacts the flavor of a coffee, and yes, sample roasting will impact the flavor. Even the coffee’s origin characteristics would not be revealed (and in some cases hidden) without the application of roasting to arrange the molecular sensory building blocks into the flavors expressed in the cup.

So, whilst the SCA cupping form allows me to fully explore and communicate the experience I have with the 11 different attributes of a coffee, it leaves little room for me to consider the attributes with regard to the roast profile, and equally as importantly, it leaves me no room to talk about production roasts of a particular profile with any sort of consistency or focus.

Creating a New Form for Production

As The Crown was a new project for us all, I decided to consider a new approach to recording the way that we, as a group, cup when evaluating new roast profiles, as well as the method of how to achieve the sensorial consistency I needed to record from week-to-week and roast-to-roast.

If I were to ever write a book about roasting, It would have one chapter and one line. Under the heading ‘Roasting’, a line would read, ‘If you can’t roast, you can’t cup’. The end. It is not only advisable that the roaster – head and production – cup the coffee they roast. Is there a need to cup every single batch of consistently roasted coffee, probably not, but I have, do and will continue to do so. Thoroughly cupping each, or a selection at least, of your daily roasts, is invaluable. Those with less experience will become comfortable with the industry’s standard evaluation procedure. And through learning to cup and evaluate each coffee’s, and indeed, roast’s attributes, a roaster learns to establish a mental map from points in the roast process to attributes on the nose and in the cup. An incredibly efficient feedback loop.

I wasn’t going to reinvent the wheel, and I knew this. Many storied companies have developed their own sensorially analytical systems for evaluating roasts, including, what was then The Roaster’s Guild, who came up with the Roast Evaluation Form (a mite too close to the original SCA form, in my humble opinion), and Intelligentsia, my colleague Chris Kornman’s old steed. Chris actually helped to develop and test their cupping form, which was one of the resources I considered when I started exploring this endeavor.

One of the most gratifying aspects of the careers of most professionals working in specialty coffee is who we get to work with. For the most part, we love, respect and are constantly challenged by our counterparts to do better on behalf of the producing-side of the coffee supply chain and be more honest and consistent in our approach to evaluating coffee quality. I am privileged to say that I work beside such colleagues. At no time in this process was I alone, either with my thoughts or questions.

Cupping for all of us is never optional, in fact, the process of cupping at The Crown is a regular and protocol-driven affair. And while the SCA form has sufficed for the needs of analyzing Crown Jewels or for considering landed samples that we may want to feature at The Crown, I was never satisfied with the feedback that I was getting, or giving, during our Production Roast cuppings.

It should be noted that I don’t believe that anyone has come up with a better form to accompany the 100-point arabica grading system than the SCA. Hand on my heart, I love that damn form! I had no intention of tinkering with that format, but I did need to develop a tool that enabled me to quickly cup coffees for production, recognize any faults not immediately noted by roast color, RoC differences, or any other major markers for roast goals.

I looked over the old Roast Evaluation Form, and the Intelligentsia form, but in the end, I realized that neither was either far enough from the SCA cupping form or didn’t quite do the job of recording the difference between cupping for discrimination over description. What I wanted was a form and method that would support a repeatable, consistent collection of data that we could use in real-time.

We started to create a new form, one that was heavily weighted towards evaluation and decision-making (discriminatory) and less towards sensorial demarcation (descriptive). We chose to focus on 5 attribute categories: sweetness, acidity, body, flavor, and aftertaste, divorcing them completely from any reference to the 100-point scale, and creating a simple scale. Each category has the potential to score positive, negative, or neutral in qualitative relation to the standard.

It was over this period of time that I realized the task that we, as cuppers were trying to do. We were trying to remember the exact nuances of each attribute of the coffees we were evaluating. In essence, what we were attempting when cupping a different roast (same profile) of that coffee, was to recall, with perfect organoleptic snapshot memory, a coffee we tasted 7 days prior. This process was, of course, highly subjective and extremely fallible.

Introducing a Control Sample

Thus, I promptly invested in a vacuum sealer. The last piece of the puzzle I had been working on was about to fall into place. But, perhaps a little context should be injected here; I am overly familiar with two types of cupping (on a regular basis), full SCA descriptive cupping and quick and dirty triangulations. For those unfamiliar with the method, triangulations are what I like to call the Sesame Street of cuppings. This cupping is best performed with limited visual information, usually, in a red-lit room. Unlike a sample cupping, a triad of cups is laid out in a line (any other format would push subconscious biases to the fore), in the cups are 2 coffees. Therefore, there are two cups with the same coffee and one with a different coffee. By cupping all three bowls, the cupper, using only their aroma and taste evaluations, should be able to tell you which is the odd one out. Or not. Depending on the reasons for triangulation cupping and the desired outcome, either one of these results is a good one.

I have used the above technique for several reasons, usually to double-check taste differences if any of the previously-set roast goals were not achieved (time, end temperature, color differences). As the cupping was set up blind, the results are as objective as possible. I had previously used this method with a binary notion of acceptance or rejection of failed batches. What I realized now, was that I could use this same method of cupping, slightly adjusted, to fully make use of the form that we were creating.

The present of the standard, a reference roast, is the keystone of the analysis. With a standard/control on the table that had been vacuum-sealed, dated and preserved from the prior week, we could now forget all notions of ‘remembering’ what we had cupped previously, as it was there, in front of us. Using the new form, we could quickly evaluate whether or not a roast had changed significantly or not at all from the previous roast, and, in real-time. In addition to this, cupping blind and making decisions with quick descriptors of any attribute discrepancies has not only served to tighten our roasting and production operations, but it has also helped shave almost 45 minutes off the time it takes to cup through 10 – 20 roasts, discuss and make decisions regarding those batches. It has also allowed us to recognize, at a glance, if the overall score of the roast was a net positive, neutral, or negative.

Using the Rating System

Here’s our score system breakdown, for each category:
-3 = not serviceable
-2 = service questionable
-1 = inferior but acceptable for service
0 = on spec, imperceptible differences from standard
+1 = better than spec, pretty cool
+2 = consider as new standard
+3 = standard-worthy

Across the cupping panel, if a coffee averages -2 or below in any single category or below -5 overall, it’s flagged for further review. Conversely, +2 in a category and/or +5 overall lifts the coffee into “new standard” territory, as we constantly seek to improve the quality of our work.

Thoughts and comments are always welcome, but in the meantime, Happy Cupping!

Download the Cupping Form PDF