The saying goes “Ask no questions, hear no lies”, but in today’s world of big data, we might be answering questions without realising that they’re being asked and the opportunistic world of e-commerce is providing us with solutions and suggestions that are hard to resist – and just about impossible to ignore. Can wine capitalise on this?
Big data is all about analysing and interpreting the wealth of online information. No company can afford to ignore the importance of data-driven decision making in staying up to date with consumer trends and preferences. Obviously this is also true for wine and for both wine enthusiasts and those who are not confident about their knowledge or preferences, there is a wealth of online wine options – from personalised apps to online information in the form of blogs and vlogs to algorithms trying to predict your wine choices. But wine can be such a personal thing and your experience of it can so easily be affected depending on where you drink it, with whom you enjoy it, what you eat with it or what temperature it is served at – to name but a few things. Is it really possible to predict wine preferences with algorithms?
It is, according to Becky Rodriguez, the communications manager for Château Teyssier in Saint Emilion. She compares algorithms with the questions a sommelier would ask: “There are absolutely certain questions you can ask to help pin people’s preferences, all of which an algorithm could do…”. But not everyone agrees with this. Quoted in the same article on vinepair.com sommelier Petrullo Campbell explains: “Selecting wine for someone’s palate is not black and white… People taste things differently. One person can describe a wine as dry, crisp, and tart while someone else can describe the same wine as fruity. It happens all of the time. Interpreting taste is so personal. That’s why an algorithm can’t work.”
No algorithm is perfect, of course, but whether you indicate preference in a personalised questionnaire or whether your past online behaviour is used as measure, I do think there can be some accuracy to wine predictions through algorithms. Being an expert with very fine definitions can make you an exception, I guess, as is when you have very limited knowledge or a very vague reference.
Some opportunities are quite straightforward. Visit Bright Cellars, complete a quiz and have some wines suggested for you. It is almost like a personalised wine club. On other sites, such as the well-known Vivino, you can ask for recommendations based on your preferences – white wine that goes with seafood, costs less than R250 and comes from New Zealand, for instance. To make sure your wines are available for these options, you can of course list or advertise or pay for exposure and recommendations, but data’s intricacy holds much more opportunity for the wine industry.
Imagine your summer wine ad pops up when tourists are searching online for a beach holiday? Imagine when someone adds steak cubes and mushrooms to their cart, your Cabernet Sauvignon is suggested as the pairing partner for their basket? Based on data available from both the current search and earlier online behaviour, such wine recommendations are completely doable and is already being done. It is up to you to make sure your wines have an online presence and tick the online boxes required, to be the suggestion that pops up.
It is a complicated science and most of us need an expert to make the most of the opportunities that data bring, but I think it will be a wise investment. Many purchasing decisions are based on online and peer recommendations, they are prompted by the all-knowing big brother that is big data and while fashion and tourism and many other industries are making the most of this wealth of information, wine can’t afford to stay behind.