Tuesday, February 21, 2012

Development of A Framework for the Assessment of Wine Quality

My current objective is to develop of a framework for assessing wine quality and that effort was launched with yesterday's post on the definition and high-level scope of wine quality.  I herein examine three efforts which seek to define the elements of wine quality and then use the learnings resulting from these explorations to construct a broad-based framework for the assessment of wine quality.

The first study considered was Stephen Charters study titled The Intrinsic Dimensions of Wine Quality: An Exploratory Investigation.  In this study the author conducted interviews and focus groups with 105 participants and, as a result of that effort, developed the quality dimensions contained in the figure below.

The Dimensions of Wine Quality (Source: Figure 1 of The Intrinsic Dimensions of Wine Quality)

The study did uncover some extrinsic wine quality factors (appellation systems, classification systems) but the focus was on development of intrinsic factors -- those identified in the glass when the wine is consumed.  It should be noted that this study was conducted in Australia using Australian participants and Australian wines.

I have a number of concerns with this study.  In my earlier post on defining the wine quality space, I indicated that customers with objective knowledge use objective cues to define quality.  In this study, the author used 105 individuals (60 consumers, with the remainder divided between wine producers and other wine industry players) with the consumer participants categorized as low-involvement (24), medium-involvement (25), and high-involvement (11).  With this obvious disparity in objective knowledge, the likelihood of getting solid input on intrinsic factors is low.  And the results bear this out.  As can be seen in the figure, a number of the components appear to me to be non-objective (drinkability and pleasure, for example) and most lack a sense of dimension (How do you measure drinkability?).  So, while the model is good for directionality, the questions raised render it unsuitable for our purposes.

The second model explored was Richard Leahy's Components of Wine Quality which appeared in a December 19, 2011 post on crushpadwine.com/blog.  Leahy divides wine quality factors into two camps: primary (sensory) parameters of wine quality and secondary elements of wine quality.  In that the scenario he describes entails judging a large group of wines, I will assume that these are intrinsic wine quality factors.  The components are, according to Leahy:

Primary Elements                  Secondary Elements
Visual                                    Balance
Aroma                                   Intensity
Taste                                     Length
Texture

My issues with this list is that: (i) it only considers intrinsic quality elements; (ii) it is presented at too high a level; and (iii) no potential values are presented for the identified elements.  Also, no material is presented which allows one to assess whether a wine is of low or high quality.

The final model evaluated was the WIne and Spirits Education Trust (WSET) Systematic Approach to Tasting Wine.  A copy of the tasting sheet, showing the categories, and potential values associated with each category, is presented below.


A few points of note: (i) This mechanism is clearly targeted at the objective individual; (ii) it clearly identifies each of the categories that exists in its model and provides potential values for each; and (iii) it posits that an assessment of quality is a journey rather than a eureka moment.  In the case where a product is being assessed for quality, the metrics associated with each parameter is provided  so thatthe customer can assess that parameter against his or her requirements.  For example, in the case of a car, the parameter miles per gallon will come with a number, 26 let's say.  In the case of wine quality assessment using this tool, we know the parameter but the value or metric associated with that parameter has to be teased out through tasting; metrics cannot be assigned to flavors or flavor intensity until the wine is tasted. For someone who is appropriately trained in the method, a quality bottle of wine would have the following characteristics:
  • Appropriate quality
  • Intensity and color befitting its age and variety
  • Clean nose
  • Intensity, development, and aroma characteristics befitting its age and variety, as perceived by the taster
  • Sweetness level appropriate for the wine style
  • Balance between variety, tannin, and alcohol
  • Body appropriate for the varietal and wine style as perceived by the taster
  • Medium-to-pronounced flavor intensity
  • Flavor characteristics that are appropriate to the variety, style, and age of the wine.
Based on how well the taster feels that the wine addresses the above characteristics, it can be rated as having poor, acceptable, good, very good, or outstanding quality.

The WSET approach is comprehensive and allows an objective taster to arrive at a quality conclusion about a specific bottle of wine at a specific time.  It is conceivable that someone else tasting that same bottle may arrive at a different conclusion but that illustrates human differences (number of taste buds, for example) rather than being an inherent weakness in the approach.  The shortcoming of this approach is that it does not take account of the vast majority of consumers who will never use it.  That is to say, its lack of extrinsic quality components renders it unusable by the wine drinker who is equipped with subjective knowledge.

Let's stop at this point and take stock of the situation.  I have shown that the quality of a wine can be derived using the WSET Systematic Approach to Tasting Wine.  That tool is only suitable, however, for individuals with pertinent, objective knowledge, leaving the large mass of the wine drinking public to its own devices as it relates to assessing the quality of wine.  And assessing the quality of wine is not a parlor game; it is more often than not a precursor to a purchase decision.  A comprehensive wine quality assessment has to address the unaddressed issue in the WSET tool and provide a mechanism for the assessment of wine quality by those with a subjective bent.  We have identified a number of these subjective cues and have coalesced them with the WSET tool to create a comprehensive wine quality assessment framework.  The proposed framework is illustrated in the figure below.


The extrinsic cues can be used either singly or in combination by the customer in order to arrive at quality decisions.  For example, the customer may use the objective opinion of the wine critic as an indicator of quality and use that as a basis for a purchase decision.  Or, on the other hand, the customer may use pricing and vintage cues in order to make that decision.

In closing, I would like to note that quality in the world of wine differs from quality in most other products and services because of the variability in terroir, production methods, and, most importantly, the differences in perception that accrue to us a result of our human condition.  Wine quality, at the end of the day, is a truly personal assessment.

I will be reviewing the various elements of the wine quality assessment framework in future posts.

© Wine -- The View From Orlando

3 comments:

  1. I actually rather like the first one (mostly the conception of paradigmatic vs. gustatory elements; I agree that pleasure is kind of silly, and could arguably be taken as synonymous with quality), but it seems more like an outline of a system still in the early stages of development than a ready-to-use heuristic.

    One question I have about your framework: how would you say wine labels (as in, their aesthetics) come into play? Are they part of the brand, or could they perhaps be considered a less objective manifestation of appearance?

    As usual, a great post.

    ReplyDelete
    Replies
    1. Thanks for the compliment.

      The literature places great store on the label as a key wine selection criteria (see the following two blog posts (i) http://winetimes.co.za/at.php?url=http%3A%2F%2Fheinkoegelenberg.typepad.com%2Fblog%2F2011%2F05%2Fthe-importance-of-the-wine-label.html and (ii) http://seattlewineblogger.blogspot.com/2007/09/on-importance-of-wine-labels.html) so I will add labels to my model as an extrinsic factor.

      Delete
  2. Very good article you have on analyzing wine which I can share in our wine lessons and events.

    ReplyDelete