April 15, 2025
What Can Big Data Do for the Wine Industry?
The big data phenomenon permeates modern society. In this new revolution, decision-making relies more on data and analysis rather than intuition and experience. Businesses can solve problems in ways that were not previously possible. The wine industry is not immune to the data bug, and some companies are already taking advantage of its capabilities.
From ecommerce to analyzing the performance of wine bottle labels, the benefits of implementing a data strategy in the wine business seem endless.
A Brief History
Data collection is ancient. Early data were in the form of tally marks used to keep track of inventories and other trade operations. Later came more sophisticated methods of accounting, such as double-entry bookkeeping — debits and credits.
Unlike today, people recorded early data by hand and analyzed it using their own mental faculties. At present, it is humanly impossible to collect and analyze the plethora of data available without the help of computers and advanced technology.
Roger Magoulas, director of market research at O’Reilly Media, coined the term Big Data in 2005. We can define it as ‘extremely large data sets that may be analyzed computationally to reveal patterns, trends, and associations, especially relating to human behavior and interactions.’ It refers to data sets — structured and unstructured — that are too massive and complex for traditional data processing software to manage.
Data sets can consist of billions, even trillions, of records.
A 2013 report says 90% of the world’s data had been generated within the last two years. It has grown exponentially since then.
Data can come from internal sources, such as transactions, or external sources, such as social media. Big data is about applying the full spectrum of data available with the potential to transform a service or organization.
“The power of big data derives from collecting vast quantities of information and analyzing it in ways that humans could never achieve without computers in an attempt to perform the apparently impossible.” — Brian Clegg, Science Author
Netflix and Chill
Netflix is an example of a company that transformed itself using the power of big data. Once a DVD rental service, it moved to streaming videos online. When we watch a DVD, the information travels one way — we receive the data stored on the disc. On the other hand, the movie service Netflix now provides utilizes more two-way communication that necessarily involves managing large amounts of data.
Netflix knows who is watching what, when, and where. It uses this information to measure viewers’ interests and provide recommendations.
Netflix’s analytics go beyond views, however. The company did not commission its big hit series House of Cards on a mere hunch and seemingly good plot. Rather, Netflix analyzed its mass of customer data and determined it had a large audience who not only appreciated the dark humor, but also liked the work of director David Fincher and actor Kevin Spacey.
Traditionally, networks have produced pilots and tested them on various markets before deciding whether to fund a project for a series. However, Netflix invested $100 million up front for the first two series of House of Cards based almost entirely on data. It turned out to be a huge success.
Uber Data
Uber revolutionized mass transit using data.
On the frontend of Uber’s app, we set a pickup location, request a ride, and a vehicle is at our doorstep, usually within ten minutes, ready to take us to our destination. Uber automatically deducts payment from a bank account linked to our profile. The simplicity and convenience of Uber’s app is one of the reasons the service has become so popular.
Behind the scenes, Uber uses advanced algorithms to analyze its massive database of drivers, processing the information with überspeed to match the driver closest to you. Uber also stores data after every trip to forecast supply and demand, as well as set fares, using predictive analytics in real-time based on traffic patterns.
Netflix and Uber obviously are two major examples. Other well-known companies that utilize big data include Amazon, Google, and Facebook. However, a company need not be a corporate giant to leverage big data.
Big Data in Little Wineries
Enolytics is a company that is “committed to bringing the power of big data to the wine industry.” Co-Founder Cathy Huyghe has been writing about wine in Forbes Magazine for nearly 15 years. She developed the idea for Enolytics about four years ago when discussions around innovation, the digital economy, and business of wine led her to realize raw data was underutilized in the wine industry.
The company was founded in 2016. “The wine industry was not ready,” says Cathy. However, over the past three years, it has gained momentum largely because of Enolytics. Cathy describes it in two waves. “In the first wave, we had some early wins, big clients asking to fill blind spots around consumer data in wine.” The data was already there. They just needed the skill of data people to work with it. They would buy data from partners. Though, what’s happened more recently is small wineries saying, “Third-party data is good, but we have our own data we’re not utilizing.”
The second wave is using internal data to make sense of things.
According to Enolytics’ blog, internal data can include “depletions, ecommerce sales, viticultural analysis, location inventory and FOB changes.” Depletion, for example, refers to sales from distributors to retailers. It’s more than a matter of simply tracking units sold. Depletion data is about knowing where the products end up. This presents opportunities to increase sales volume as wineries can leverage their knowledge of which wines sold the best in a particular region. They can use this data to identify trends over time, visualize it for specific markets, and present it to distributors accordingly.
Wineries also can compare depletion data with ecommerce sales to improve business processes. They even can analyze viticultural data to implement sustainable water management. There are multiple ways wineries can take advantage of internal data. A major benefit of internal data, particularly for small wineries, is they already own it and don’t have to pay for it. They just need the right team to analyze it.
In Artificial Vintelligence, I cover how we can use chemistry, data science, and machine learning to predict wine quality. Tastry is a data and insights company that uses similar methods to predict what wine you will like before you try it. They like to say they “taught a computer to taste.” Combining analytical chemistry, flavor preferences, and machine learning, Tastry can track consumer preferences in not only wine, but other sensory-based products such as beer, coffee, cannabis, perfume, and food.
While studying chemistry at Cal Poly State University, and working at a custom-crush wine facility, Tastry’s founder, Katerina Axelsson, noticed a major discrepancy in the scores of two bottles of wine from the same batch. This led her to the idea of developing a more science-based approach. Thus, Tastry was born.
AI Meets Chemistry
“We looked at the chemistry [of wine] for years,” says Katerina. First, they use AI to break down the ‘flavor matrix’ of a product. Thousands of chemical compounds in wine makes this a complicated task, but a “good problem for AI to solve,” according to Katerina. Once they decode the flavor matrix, Tastry generates a ‘flavor profile’ to describe the “relationship between the chemical compounds within sensory-based products and their relevant expression on the human palate.”
What sets Tastry apart from other companies that track consumer preferences is it uses both chemistry and machine learning. After identifying a consumer’s ‘palate profile’ with a 20 second survey, its patented artificial intelligence analyzes the data to match their taste to products in its database.
Tastry employs what it calls “Singularization.” This is the process of “matching a specific individual to a specific product without any previous purchase history.” Customers also are not required to rate other products. They refer to this as a “small data” approach. It “mitigates the ‘cold start’ problem typical of most recommender systems.” As a result, consumers are 45% happier with their purchase.
Wine Retailers Are Reaping the Benefits
Tastry offers a kiosk retailers can set up on their sales floor. This is particularly beneficial for grocery stores that typically don’t have wine specialists on staff to assist customers. Tastry reports a 5–10x return on investment in the first year of retail implementation. The data and insights allow stores to optimize their product mix. They know what preferences customers will buy and can stock wines accordingly.
Not only does this benefit retailers, but distributors and manufacturers as well. Distributors can gain valuable insight on where a wine will perform best based on local, regional, national, and global market data. Manufacturers can benefit by using a data-driven approach to tailor wines to the taste preferences of intended target markets.
In a recent test, Tastry blindly tested 125 wines and predicted consumer ratings with 91.8% accuracy. It also showed how they can increase the estimated value of a wine by 208% by providing blending directions.
What’s in a Label?
Label Analytics gathers and tests data on wine labels. Its mission is to “enable producers to sell more wine.” 50–80% of buying decisions are made at the retail shelf. The average retailer stocks between 900 and 1100 wines. Naturally, wine labels play an important role in grabbing a customer’s attention. Label analytics’ job is to unwrap the science of label designs that motivates purchase decisions.
Label Analytics tells producers how likely wine shoppers are to buy their bottle off the shelf. It prepares Comprehensive Wine Shopper Response Data winemakers can use for sales pitches. Wine producers generally don’t invest much in commercial advertising. Their label is their primary source of advertisement customers see when shopping. Label Analytics replaces a lot of guesswork with data-based decision making.
$203 Billion Market
There are various ways the wine industry can implement data analytics. The foregoing is just the tip of the iceberg. Opportunities abound.
The big data and business analytics market will reach $203 billion by 2020. While the wine industry has a $220 billion impact on the U.S. economy, it is unclear how much it will invest in the analytics market. Though, one thing is clear — data is a major asset to virtually every industry. New generations of entrepreneurs are using it to revolutionize the business of wine.
“Data is a central resource. Only those who manage to extract value from their data will remain competitive.” — Dr. Sebastian Derwisch, Data Scientist
The foregoing article was originally posted on my wine blog. Some of my most popular articles have been reposted to Medium. Interested in more wine content? Click here.
