The back wine label: What information and features appeal to consumers?
By: Kathy Kelley, Jeff Hyde, Professor of Agricultural Economics, and Johan Bruwer, Professor of Wine Business and Marketing, the University of South Australia
It’s common knowledge that many consumers purchase wine based on the label design, and that the design often influences consumer perception regarding wine quality and expected price. Though several research studies have focused on front label characteristics, fewer have investigated the importance and usage of the back label. In one study, 57% of respondents reported that they read the back label during the decision process (Charters et al., 1999). Among these participants, more women (61%) reported that they read the back label when shopping compared to men surveyed (51%). In another study (Thomas and Pickering, 2003), participants rated the importance of both the front and back wine label in the purchasing decision using a 7-point scale (1: very little importance to 7: very great importance). Though statistically significantly different, ratings for the front label and back label were 4.79 and 4.60, respectively.
As the information on the back of a wine label can vary greatly from a description of where the grapes were grown to images and maps that supersede text, Penn State researchers and a faculty member from the University of South Australia wanted to learn if including the following information on the label would appeal to wine consumers:
- Three options for food pairing information: 1) no food pairing information, 2) text-only description of food pairings, and 3) images of symbols of food pairings
- How to contact and connect with the tasting room: 1) Quick Response [QR] code, 2) tear away strip, and 3) winery tasting room email and website address
- Four winery background information options: 1) information about the winery only, 2) the wine maker and winery, 3) the winery and wine region, and 4) the wine region only
We selected these elements based on what we found to be present on wine bottles sold at tasting rooms and at the Pennsylvania Wine and Spirit Stores, some of which appeared more frequently on back wine labels than others. For example, we investigated the appeal of a website and email address for a winery tasting room as consumers often contact businesses and access information about products via the Internet. We then chose to include a Quick Response (QR) code to the back label as smartphone and tablet users who scan the two-dimensional code are directed to a wineries’ website or any other webpage associated with the code. While invented in 1994, this technology is relatively new to the wine industry as reportedly being first used by a winery in 2010 (Franson, 2011), which corresponds to when we conducted the survey. Even today, the use of QR codes on back wine labels is still quite low.
We also investigated the appeal of a “tear away strip” (perforated part of the wine label that contains the name of the winery, varietal, and other information) that consumers can easily remove from the label and use to remember the wine and refer to when making future wine purchases.
Design and statistical techniques used
Between July 12 and 17, 2011, 910 consumers who were age 21 and older, who drank and purchased wine at least a few times a year, and who lived in either metropolitan Philadelphia or New York City participated in a 15-minute Internet survey. Among the survey questions, participants were asked to evaluate 16 mock labels, which contained one of the options from each of the categories of interest: food pairing, how to contact and connect with the tasting room, and winery background information. We also added the following to the mock labels so that each would be as representative as possible of what would be on a commercially available bottle of wine: fictitious brand name and bottling address, 750 ML net content, declaration of sulfites and health warning statement (Alcohol and Tobacco Tax and Trade Bureau, 2008) and fictitious UPC code.
If you calculated the number of mock labels that were possible, based on the number of different attributes we tested (3 different food pairing options x 3 how to contact and connect with the tasting room options x four winery background information options), you would come up with 36 combinations. By using a statistical technique called “conjoint analysis” (used frequently in market research), however, the number of actual combinations that consumers needed to evaluate was reduced to 16, which was a much more feasible number to evaluate.
An example that might help you understand why we use this technique can be explained by the following example: Consider an automobile manufacture that wants to introduce a new model to their SUV line. With all numerous seat types and upholsteries, interior design styles, dashboard layouts, pricing, etc. that could be offered how do they select the winning combination and where do they begin? Conjoint analysis can ultimately determine, based on scenarios that asked consumers to indicate how likely they would be to buy a vehicle with “x” seat type, “y” interior style, “z” dashboard layout, etc. This technique can then identify which feature (e.g. seat type) was the most influential in their decision to “purchase the vehicle,” as well as what seat type (e.g. x, y, or z if three different seat options were investigated) consumers preferred the most. In essence, we did the same with our wine labels. We were able to learn if food pairing in general was more appealing then contacting and connecting with the tasting room and if either of these were more important to our participants than the addition of winery background information to the label. Also, of the three pairing information options – which was the most appealing and which was the least.
It is important to note that even through participants evaluated less than half the total possible number of combinations – all food pairing, how to contact and connect with the tasting room, and winery background information options are represented and tested equally.
Our survey participants
Sixty-five percent of the 910 participants lived in the New York City metropolitan area and 35% lived in metropolitan Philadelphia. Approximately one-third of participants resided in each of the three states that comprise the two metropolitan areas: New York (36.6%), New Jersey (35.2%), and Pennsylvania (28.2%).
Most common responses to demographic questions were female (71.9%), a member of a two-adult household (54.1%), resided with at least one other adult wine drinker (69.6%), lived in a household with no children (55.9%), were age 35 to 44 (36.8%) years, had a bachelor’s degree (37.7%), with a household income between $25,000 and $75,999 (46.4%).
After responding to questions about wine consumption and purchasing (which will be presented in future blogs), participants rated each label using a seven-point Likert scale (1= very unlikely; 7= very likely) as to their likelihood to purchase the bottle of wine based on the three factors, regardless of varietal, to serve to family and/or friends. In total, we were able to collect completed responses for this part of the survey from 847 of the 910 participants.
So, what appealed most to our survey participants?
When responses from all of the 847 participants were used in the analysis, the pairing information category was found to be more important than either contact and social media information and winery background information. In fact, based on the scenario described above, 43.3% of the decision would be based on pairing information, 26.8% would be based how to contact and connect with the tasting room, and 29.6% of the decision would be based on background information.
We also analyzed the data based on how frequently a consumer purchased wine and how often they consumed the beverage. The only difference in order of importance of the three characteristics was found for the 64 participants who purchased wine at least once a week (7.6% of the 847 participants who completed the wine label evaluations). For these more frequent wine purchasers, winery background information was slightly more important (38.7% of the purchasing decision) followed by how to contact and connect with a tasting room (31.1%) and pairing information (30.2%).
The outcome for those who responded that they drank wine more than once a week (28.6% of those who completed the wine label evaluations) also known as “super core wine drinkers,” was similar to the outcomes for the entire group of participants: 39.7% of the decision would be based on pairing information as a whole, 28.0% would be based how to contact and connect with the tasting room, and 31.8% of the decision would be based on background information.
You may be wondering whether those who purchased wine at more frequent intervals also consumed wine at more frequent intervals. In other words, are the two groups represented above comprised of the same participants? A quarter, 24%, of participants both drank and purchased wine at least once a week while another 28% of participants purchased wine a few times a year – regardless of consumption frequency. In the study that was conducted in 2013, which has been published in a past blog, we did ask consumers more detailed questions about buying single bottles and bulk/case purchases (http://bit.ly/1FzZ8dA).
The “winning” label combinations
As mentioned above, in all but one instance pairing information accounted for approximately 40% of the purchasing decision. Therefore, as would be expected, participants did not rate labels that lacked pairing information favorably. Analysis showed that either a written description of what the wine paired with or images/symbols of the food items were nearly equal in their positive appeal. So, based on our study, it would prudent for wineries to consider including descriptions/symbols of food parings as their customers might value this information as our participants did.
Both the tear away strip and the QR code received negative ratings, while the email and website URL option received only “slight” positive ratings. Wineries interested in using a QR code, tear away strip, email/URL, Twitter/Facebook/Instagram, or other methods of allowing consumers to contacting/connecting with the tasting room should survey their customers to learn what they would find useful. Customers’ preferences might be based on generation, consumption frequency, and other factors. Regardless, it is always preferable to ask your customers what they prefer than to just make a guess. Also, there may be ways other than the back label to provide this information to those same customers.
Outcomes pertaining to winery background information indicated that two options were not favored among our participants: 1) when only information about the winery was provided and 2) when only information about the wine region was included on the label.
Instead, our participants found the other two options appealing: 1) information about the wine maker and the winery and 2) information about the winery and wine region (though ratings were similar – this option was slightly more appealing).
What does this all mean?
Certainly there are many components and combinations that we could have tested in our study, but since we were relying on consumers to evaluate more than one piece of information on the labels (and we included the government warning, winery address, and UPC to make the label as representative) we did not want to overwhelm our participants with even more label options. There is a recognized limit as to how many conjoint analysis combinations consumers can likely evaluate before becoming fatigued. We did not want this to become an issue in our study.
As stated previously, we tried to identify components that had been shown to be of value to consumers and either commonly used on back labels or could have potential to enhance a consumer’s interaction with the winery and tasting room.
Other factors can certainly have an impact on whether or not a consumer finds a back label appealing and useful: label paper thickness, gloss/no gloss, and color; fonts style, size, and color; as well as brand, varietal, and price (which we tried to remove from the equation), etc.. Still, most wineries spend time developing back labels, printing, and applying them to bottles. Our research provides those who are evaluating their current labels or developing their very first label with a bit of insight – a starting point for crafting a back label that could include items our participants found appealing. Marketing is a science and even if you do not have access to statistical tools like we do – you have access to the most important piece – consumers who will be purchasing your product.
We have much more data to share, from both this study and others. As with the information that has been shared in past blogs, our hope is that we can provide at least one idea to help your winery and tasting room be successful.
Charters, S., Lockshin, L. and Unwin, T. 1999. Consumer responses to wine bottle back labels. Journal of Wine Research, 10(3): 183-195.
Franson, P. 2011. Wineries connect with QR codes. Wine Business Monthly, 18(8): 42, 45-48.
Thomas, A. and Pickering, G. 2003. The importance of wine label information. International Journal of Wine Business Research, 15(2): 58-74.
United States Alcohol and Tobacco Tax and Trade Bureau, 2008. What you should know about grape wine labels. U.S. Department of the Treasury, available at: http://www.ttb.gov/pdf/brochures/p51901.pdf (accessed 24 August 2014).
- Customer service
- Hard Cider
- Sensory Training
- Students & Projects
- Team Members
- Wine Marketing
- Wine Marketing & Research Board