Home / Commerce/Gestion / “We have created the most advanced commercial search engine ever built for wine”

“We have created the most advanced commercial search engine ever built for wine”

The huge surge in online wine sales since the start of the pandemic has been widely documented, but concurrently with its volume increase, the channel has witnessed changing consumer habits and demand. Recent research by the IWSR has shown that consumers are increasingly willing to experiment, trade up and find out more about the wines they are buying. The newly-launched Pix platform intends to capitalise on the trend and draw on industries outside wine to enhance the user experience of online wine buying. We asked its founder and CEO Paul Mabray to explain more.
By Sharon Nagel November 08, 2021
“We have created the most advanced commercial search engine ever built for wine”
Pix CEO and founder, Paul Mabray, looked outside wine for inspiration to create the new platform with its 250k+ wines - crédit photo : Pix

Pix has announced that it will be launching a new wine discovery platform. Can you tell us more?

ix is already here. We’re in Open Beta and anyone can access the platform by logging on with any social sign-on today. It’s taken us 11 months to get from the business plan, though to funding, building the team, and now into Open Beta. 


Why will this not just be an umpteenth wine marketing platform – what is its USP?

There’s never been a platform built by a team like this. Seriously. This is a once-in-a-generation team of top-rank wine experts, globally-renowned communicators, old-school bag carriers from the trade, and tech experts. We’ve even got leading voices in user experience and behavioural science. We’ve never seen this sort of expertise brought to bear on a wine marketing platform before.

That creates one huge difference between Pix and really everyone else. We’re all experienced enough to know we don’t know all the answers. We didn’t design the platform. Our users did in our closed beta. We tested concepts and adopted the ones that worked, and dropped the ones that didn’t. That’s a fundamentally different approach from saying “I’ve got the answer to wine search and matching” and building back from that. Crucially, we looked outside wine for inspiration, to platforms that connect users with books, podcasts, movies, music… We spoke to the people who created those platforms to understand how they tackled discovery with a long-tail product like wine.

To be honest, we started from a very different premise to other platforms. We’re here to enhance a business we love, wine, not “disrupt” it. We have created the most advanced commercial search engine ever built for wine. But we also know that it needs to be enriched by human experts, and complemented by a highly creative and talented team to give you a compelling reason to come back again and again and find new bottles to fall in love with. 


What will its key functionalities be?

First picture yourself in a wine store. The manager says, “Hi, welcome, how can I help you?” and you reply “I’m looking for…” You’ll see that’s what it says in our search bar. “I’m looking for…”. That’s the open question Pix is here to answer. As such we are building an advanced wine search engine to answer those questions. Then we have compelling content to tell you the stories behind wines by some of the most compelling voices in wine. We also take visitors on tour of regions and grapes, thanks to our collections offers. That’s only the start of what the collections are. 


The criteria for selection on the beta version seem pretty conventional – vintage, region etc Doesn’t this limit the platform to fairly knowledgeable wine drinkers?

We’re here for wine lovers whether they’re looking for Bota Box or Petrus. What we want to do is guide consumers to whatever they personally consider a great bottle, regardless of their background in wine. And, as we have discovered, even novice drinkers love discovery. Our thumbnail descriptions were inspired by the metatags you find in Netflix movie listings. We developed the idea through some testing – and they turned out to be a revelation. Texture words really appeal to people. Now there are three key terms that give a “thumbnail” idea about the wine. The first is a feel word, like rich, elegant or fresh. Then comes a flavor word, like currant, cherry, or spice, to give people a general idea of the taste. Finally, a story word, to help people visualize how they will use the wine. “Dinner with friends”. “BBW Wine”. We also know that people are becoming deeply interested in who makes the wine and how, so the story terms might include “female winemaker”, or “BIPOC winemaker.” 


Will producers/marketers be able to sell wines through the platform, or will it always be a matchmaking concept only?

The platform is built for wine sellers of all types to sell wine. Whether that is by adding better information about their wines, ensuring their stores are connected, and, if they want to get more focus on their brand, they can buy and boost key words. That allows them to sell more wine both directly and/or indirectly.


The platform is being touted as the second-biggest wine platform in the world at launch. How many listings are we potentially talking about?

Second biggest platform with wines for sale: over 250K wines.


Which countries have shown most interest in the platform so far?

We’ve had enquiries from many countries. This week alone, we had enquiries from Turkey, Sweden and Singapore. We’re launching in the US which is the world’s biggest wine market by value and will soon launch in the UK and Canada. Once we have those countries up and running, we will expand further.


Does it depend on how tech savvy they are?

We are a platform for all wine consumers so we expect anyone looking to sell wine in the US and UK will be interested in working with us. The team and I are talking to winemakers around the world and we’re finding tech savvy producers everywhere. People recognize it’s the way forward. They want help from someone like Pix to support them.


You intend to offer the opportunity for producers to pay to promote their products to specific audiences. How will this work and is this the platform’s only source of revenue?

You can buy keywords and focus them by market and channel. Like Google, it is a very effective methodology. Wineries and retailers can buy words and phrases that help them sell wine. Whether that is “Cabernet”, “Caymus Cabernet 2017”, “Helen Turley”, or “best wine for date night” there are infinite options and combinations of words, geography, and channels to target. Just like Google, Instacart, Amazon, Etsy and other companies that leverage keyword purchasing, it is successful because it is effective. Buying keyword placement will be our primary source of revenue and will almost exclusively be our focus for the first 12 months.


How about straightforward advertising?

No. We don’t want any conflicts of interest.


Can you describe in layman’s terms the technology used to ‘match make’ between wine drinkers and their ideal wines?

We believe that our job is to combine human and machine, to help teach a machine how to understand wine, how to organize wine, and how to emulate the real world experience of walking into a store. As such we have a Master of Wine helping to teach the engine about all the elements of wine, how wines are connected, and how to organize those wines. He’s basically putting it through advanced wine education.


Are wine drinkers really so predictable that algorithms can tell them what they should be drinking?

No. Algorithms currently are garbage for predicting taste. All those ridiculous taste tests are junk science. But we can understand consumer behavior by creating an effective tool that helps consumers at point of purchase. That way we better understand their taste preference. If you only scan, search and click on unoaked chardonnay, there’s a good chance that if we show you Rombauer it will be a bad experience. If you only view deep, rich reds, we won’t recommend Beaujolais. That being said, the platform should, via content and merchandising (both programmatic and human), introduce consumers to new wines and experiences they may love.


Are wine drinkers such a globally homogenous group that their tastes are the same the world over?  

We don’t think that the wine industry is homogenous. We don’t even think a single drinker is uniform in their consumption behaviour. As such we’ll make fine adjustments, down to the individual consumer and their unique preferences. Think about it: You might drink rosé in the summer. Chianti in the winter, Prosecco on Wednesday, and Champagne on Sunday. You’ve been part of four different groups depending on your situation. This is true of wine lovers everywhere, though in some markets they may have preferences weighted more towards specific styles.


If Covid hadn’t accelerated the migration to online wine buying, would the platform have existed?

No. The Black Swan event of Covid catalyzed the awakening of wine online across all tiers, from producers and wholesalers/importers to retailers, as well as consumers.


How do you see the future of online wine retailing?

The growth of wine online will continue to accelerate. This was an inevitability that has been accelerated by COVID. I expect we’ll see enormous growth and innovation in the channel over the next 10 years.


Do you have other plans for using AI in the wine arena in the near future?

We believe in using the right tools for the job, not inserting technology for technology’s sake. We have an amazing team of people, who generate great ideas, and then use the most appropriate technology to realize them. We use AI for our matching engine. We will use AI for our dynamic ad engine. And we will use AI to help us build better wine recommendations. But unless there is a job to be done that helps truly serve the consumer, we won’t add technology beyond what creates value for the consumer.


All Comments (0)

No comment to this article.
Vitisphere 2022 - Tout droit rserv