In the specialty coffee industry, it’s widely believed that screen sizes can be indicators of quality. While large, uniform beans can make roasted coffee more visually appealing, can the size of a bean bring out certain flavor attributes? Mike Romagnino shares findings from his experiment that set out to determine if screen size has a direct impact on flavor.

Comparing Coffee by Screen Sizes

Most coffees are sorted by size at origin while being prepared for bagging and export at a facility called a Dry Mill. To achieve uniform bean size, the coffee is passed over a vibrating table filled with holes of varying sizes. This process sorts the beans by size to meet the specifications of a client’s contract. Screen sizes are measured in terms of 1/64th of an inch. For example, a size 17 screen bean would measure 17/64th of an inch. Most specialty coffee falls between 15 and 18, roughly a quarter of an inch. To learn how different origins classify screen size, read this Coffee 101 blog post.

One of the most common examples of size separation is found in Colombian coffees. In Colombia, coffees are either classified as “Excelso,” meaning screen 15+, or as “Supremo,” screen 17+. The common assumption in the trade is that the larger screen size tastes better than the smaller.

We set out to answer this question: Say that you have two coffees from the same farm that were picked and processed at the same time. If their only differentiating factor is the size of the bean, would the coffees perform the same way on the cupping table and offer the same flavor profile?

Sorting & Roasting Different Coffee Sizes

To test this theory, we used a coffee graded 15+, meaning that the beans inside the bag ranged from screens 15 through 20. Then, we sorted the coffee into screen 15/16 sized beans and 17/18 ones. Here’s a breakdown by percentage of what was screened on average from each sample:

2% – Screen 13/14
19% – Screen 15/16
66% – Screen 17/18
13% – Screen 19/20

Next, we roasted the coffee in 1lb. batches on our Stronghold S7 due to its automation and preprogrammed profiles. Our goal was to eliminate the human element from the roasting process and ensure consistency from roast to roast.

In order to test the potential for varying density in larger vs. smaller beans, we utilized different roast levels. We selected two preprogrammed roasts for each screen size: one at a light/cupping style of roast and one for a more developed, medium roast.

Cupping & Evaluation

A day later, we were ready to put the roasts up on the cupping table for evaluation. However, we chose to set up a blind cupping for this exercise. There were two sets on the table with 4 cups per set. One set compared the automated light roasts from the Stronghold while the other compared the medium roasts.

In the first set, the light roasts were clean and crisp with light body and dominating notes of citrus. The medium roasts also presented a clean cup but showed heavier body with notes of milk chocolate, plum, and orange.

After cupping through the light and medium samples, we could not determine a difference in the flavor profile from cup to cup in either set. The flavor profile for the light roasted coffees was identical in every cup as was the flavor profile for the medium roasted ones.

Final Thoughts

Performing this roasting and tasting exercise was both worthwhile and informative. From my experience, there is a perception that larger coffee beans taste better because of their size. Additionally, people sometimes assume they taste better because of their visual appeal. This is evident when buying and selling offerings like Kenya AA and Kenya AB or Colombia Supremo and Colombia Excelso. In some whole bean applications, we know that large, uniform beans can present better and warrant a small premium. However, the size will likely not improve the overall cup score.

At the end of our experiment, we found that screen size did not seem to matter when roasting & cupping this specific offering. While we cannot use this example to make broad assumptions about how coffee screen size relates to quality and flavor, it’s fair to say that different screen sizes in a single batch may not ultimately affect flavor profile.