Wine o’clock? How a clever algorithm called Bacchus takes the guesswork out of what to drink

Avatar photo

Fiona Fraser

This month, New Zealand subscription service WineFriend turns five. What started as a simple plan to deliver tasty wines to customers based on what they said they liked, WineFriend has evolved into a data-rich platform that fine-tunes and personalises selections with each bottle you try.

When Debbie Sutton dreamt up WineFriend, it was as a solution to the problems she’d encountered in her previous life as a boutique wine label founder. “As small producer it’s so so hard to make headway,” says Sutton. “The big retailers and the wine buyers hold all the power, and you have none.  And then there are the consumers, who often get into a rut of buying the same wines week after week or sticking to a rotation of three to four brands. They’re trained, by the supermarket model, to only buy on promotion.”

Wine is a huge category in the New Zealand supermarket space. “There are roughly 48,000 items in the store and 1500, or more, will be bottles of wine. It’s the number two category, after breakfast cereal.”

Sutton thought she could do better, offering choice, and a space for smaller producers to grow awareness and build loyalty. So, she went to My Food Bag, the (at that time brand new) meal kit subscription service and asked who had developed their proprietary website. She wanted one, only for wine.

And she wanted it to be really clever. Enter Bacchus – named for the Greek god of wine but also because it’s cute wordplay on the site’s back end. And part of what makes Bacchus so great is his algorithm that tailors each order to the customer’s taste in wine.

WineFriend’s highly personalised model delivers wine based on preferences.

When you join WineFriend, you take an initial online “taste test” which asks questions that delve deep into your likes and dislikes – things such as how you take your coffee, and if you add salt, hot sauce or chutney to your meals. The answers give Bacchus a baseline, and helps sort out a wine order for you. But it’s your response to the wines in that order that inform the next order, and that one the next and so on.

“We are really the only wine retailer that encourages two-way communication with our customers,” says Sutton. “Customers can quickly review the wines they’ve tasted and at midnight every night we get an email digest of everything that comes in from the past 24 hours. That’s how we can really measure the pulse of our customers.”

Bacchus knows which wines are in play at any given time – 75 of them, chosen by wine writer and Chief Tasting Officer Yvonne Lorkin. ”Yvonne and our wine buying team choose a mix of varietals and wine brands and pop them all into categories – it’s like a big library of wine! And once your payment flicks through, the algorithm then picks the best wines for your tastes. It takes information such as what you’ve really loved before and if that particular wine is back in stock, it might come back to you in your order with a little love heart sticker on it. It’s a really nice way of showing people we are listening to them.”

With Bacchus holding almost 90,000 wine reviews from customers, Sutton and her team are able to gather insights on what’s happening in the wine-drinking world too – what’s hot, what’s trending, and what’s tired. “So we know right now that rose is big all year round, and prosecco has really taken off. But our stats also tell us that our customers are a lot more adventurous than regular supermarket wine buyers. They are willing to try new things.”

And that’s what’s really exciting to Sutton. “There is no other wine subscription service in New Zealand doing quite what we do. Nobody else can back up the claim of personalisation. We have an average of 27 months’ customer retention which is great for a subscription-based business. People stay with us because they love the wines and appreciate the customer service. And we really love hearing from them.” In fact, a favourite activity in the WineFriend office is reading customer feedback out loud to each other. “Sometimes, we’ll pop what we call a ‘wild card wine’ in with the order – something that’s a little way out of the customer’s known comfort zone. So many love the wild card, though, and reach out wanting more!”

Sutton says as with any tech based business, she’s always looking at what’s coming over the horizon.  Bacchus has had three major iterations so far, and there will be more. “If we’re not working on enhancements, we’re thinking about them. With software you’re never finished. “

But right now, it’s time to stop, and to celebrate turning five. “We’ll be having a drink for sure.”

Avatar photo

Fiona Fraser

Fiona Fraser spent 18 years as a journalist and editor before founding Contentment Agency, her content and public relations business. From first getting behind a radio sound desk as a teenager, to thrashing podcasts as an adult, she appreciates the myriad ways tech can enhance communication and connection.

Data, AI, BI & ML

See Profiles

Artificial Intelligence and Machine Learning are the terms of computer science. Artificial Intelligence : The word Artificial Intelligence comprises of two words “Artificial” and “Intelligence”. Artificial refers to something which is made by human or non natural thing and Intelligence means ability to understand or think. There is a misconception that Artificial Intelligence is a system, but it is not a system. AI is implemented in the system. There can be so many definition of AI, one definition can be “It is the study of how to train the computers so that computers can do things which at present human can do better.” Therefore it is an intelligence where we want to add all the capabilities to machine that human contain. Machine Learning : Machine Learning is the learning in which machine can learn by its own without being explicitly programmed. It is an application of AI that provide system the ability to automatically learn and improve from experience. Here we can generate a program by integrating input and output of that program. One of the simple definition of the Machine Learning is “Machine Learning is said to learn from experience E w.r.t some class of task T and a performance measure P if learners performance at the task in the class as measured by P improves with experiences.”

You might also like

[ajax_load_more id="9462509724" container_type="div" post_type="post" posts_per_page="6" post__in="" pause="false" placeholder="true" scroll="false" button_loading_label="Loading" button_done_label="No results" no_results_text="No results"]

Our Vendors

Subscribe to
WineFriend's Algorithm Selections | Umbrellar Connect

Get the latest news content in your inbox each week