Coffee Cuality™ – A New Method for the Evaluation of Coffee Sensory Quality
There currently are a number of ways to assess coffee sensory quality, most of which produce an overall quality score on a 100-point scale. We believe this is useful and necessary information.
Even though this quality score is often arrived at by adding sub-components that refer to the different modalities involved in coffee sensory perception (i.e., appearance, aroma, taste, flavor, mouthfeel), it is difficult to find a justification of the overall quality score, other than the fact that a high scoring coffee will receive high scores for all sensory subcomponents and a low scoring coffee will receive low scores for them. The interpretation thus becomes complicated and the overall quality score is not that actionable in terms of looking for specific reasons for the high or low score. And practically speaking, this means that a judge typically comes up with the overall quality score, before then scoring the subcomponents so that they add up to the intended score out of 100 points.
In consumer research
we have developed means to justify, explain and then act on the overall acceptance score a consumer gives to a product. A typical scorecard in the consumer tests that we conduct starts with a so-called hedonic rating of overall liking on the ubiquitous 9-point hedonic scale that goes from 1=’dislike extremely’ to 9=’like extremely’, with 5=’neither like nor dislike’. By itself, this hedonic rating only tells us if the consumer likes or dislikes the product, and how much, but it does not begin to tell us why that is. Somewhat the same challenge as the one we face with experts rating the quality of coffee on a 100-point scale.
With hedonic ratings from consumers, we usually begin by investigating preference segmentation through a set of techniques called preference mapping. Indeed, consumers typically have different likes and dislikes for a beverage like coffee. Preference mapping has three purposes:
Segment
How many groups of consumers exist with different likes and dislikes?
Drivers of Liking
Investigate which sensory attributes (positively or negatively) drive liking by each preference segment.
Characterize
Identify the unique demographic, psychographic and usage characteristics of the uncovered preference segments, if any
Certainly, after we ask the overall liking question, we usually ask how the consumer likes the appearance, the flavor (aroma, taste & heat), and the texture/mouthfeel of the product. Theoretically, that should help us understand which of these sensory modalities drove the overall liking score. But in reality, we always find that the greatest correlation with overall liking is liking of flavor, then liking of texture/mouthfeel and finally, liking of appearance. Regardless of whether we are evaluating tomatoes, crackers, ice cream or even chewing gum. So we are left frustrated with our incapacity to explain the degree of liking for the product.
We have therefore designed and incorporated two additional strategies in consumer research to figure it out.
1) Just-about-right scaling
This strategy involves asking the consumer how (s)he feels about the level of key sensory attributes in the product. Take saltiness or crunchiness, for example. Is it too low, too high, or just right? We can examine the scores and see whether a suboptimal hedonic rating was driven by an inadequate saltiness, by looking at the frequency of consumers choosing a rating other than just right. And we can conduct a Penalty Analysis to quantify how much is lost on that 9-point scale when we depart from ideality. Do we lose half a point if the product is not salty enough or do we lose a whole point?
2) Check-all-that-apply
This method has the consumer select from a list of descriptors of the sensory quality of the product – a more analytical task usually reserved to trained experts conducting descriptive analysis. Liking by consumers can then be related to this analytical information to identify so-called drivers of liking in the sensory profile of the product.