Confused about which wine to buy? Maybe now there’s an answer. Katerina Axelsson and her company, Tastry, have been teaching machines how to taste wine like a human. Answer a few questions on a quiz and this AI program is able to recommend the perfect wine for you!
The start of Tastry
Katerina Axelsson started her entrepreneurial journey at Cal Poly University. She paid her way through college by working at a custom-crush facility, where people interested in making wine could get access to equipment. There, she wore many hats and spent hours in the lab. She had the freedom to run her own experiments and “act like a mad scientist” after work.
While working in the lab, Axelsson noticed that one batch of wine distributed with two different labels received drastically different industry scores. There had to be an objective way to grade wine. Over the next three years, she began a journey to teach a computer how to taste.
“Tastry was one of the inventions that came out of my many experiments. The inspiration came to me when I noticed that there wasn’t much data behind the process. There was a lot of intuition involved.”
The experiment
The same wine can be sold under many different labels. As a wine consumer, you could be paying 50 percent more for the same product. The primary difference could be as little as a label or the marketing.
Before Tastry was ever a company, Axelsson went down a two-year rabbit hole. She read research papers on sensory science, talked to major sensory science companies and tried to understand how people perceive flavor.
“What kind of science is there for making products? People spend hundreds of thousands of dollars on focus groups, and it doesn’t seem to be improving things. Why is that? I had all these questions.”
Her success story
At Tastry, the next step was to prove that the quiz works for business. By using Tastry’s solution, on average, customers scored products 45 percent higher and retailers saw a 12 percent increase in wine sales.
“Where it started to get interesting was when we had deployed in enough locations across the US, we can see trends in consumer pallets in the data and use it against the inventory,” said Axelsson. “We can see a direct relationship with how high the pallet matches at any given store and how well they’re doing in wine sales.”
About Tastry
Tastry set up a lab using this unique method and tested the majority of the U.S. wine market. Now, Tastry needed to validate that it could use this chemistry to predict what consumers were going to like.
The first product it ever launched was a wine recommender engine. It’s a whimsical-looking quiz—you answer roughly 10 questions on how you like your coffee, licorice, chocolate and other foods. It then recommends the top wine products in any given store.
The quiz’s underlying foundation is chemistry. To accomplish this, Tastry poured thousands of wines to thousands of people to have them blind rate the wines and help train the algorithms. By using that data, Tastry can predict how consumers are going to react to the product. Tastry is a sensory sciences company that uses chemistry to make products and recommend them.
“Once we finish that process, Tastry doesn’t need a historical data set of how many products you’ve ever tasted to know what wine you will like. All we need for now is 10 questions that have nothing to do with wine,” said Axelsson. “We know exactly how to recommend something that you’re going to like before your first sip!”
Advice for other CEOs
Axelsson started Tastry with the mindset of a scientist. This worked when the business was small, but as the company grew, her viewpoint needed to change. Thus, Axelsson began her mental transition from scientist to CEO.
“It’s very easy to have a great company culture if you start out with the right people in the first place. I started with chemistry and inventing, but then I hired experts and met people with master’s degrees to take it over and improve on it. Then, I moved on to something else and then hired people smarter than me to take over that role.”
Adapting to your company’s current situation is a lesson that can be learned from Axelsson. Transitioning to a different role in your own business can be a daunting task. Recognize that change can be beneficial to the culture, company and yourself. If everyone is moving forward together, then success takes care of itself.
Add Comment