Non-Food & Beverage related articles are being moved to Medium. Find those posts about blockchain and AI at medium.com/@dbchain

Is the wine industry against the personalization provided by AI?

Why do some wine professionals resist the use of AI?

3/3. This article is the third of a three-part series dedicated to AI and the wine business. In the first article, I review the AI techniques that enable consumers to become connoisseurs. In the second article, I reflect on the marketing techniques of the wine reviewers. In the last article, I suggest ways to use AI based on winemaker profiles.

I had the pleasure of interviewing leading players in the wine industry to write this article. Thank you to the interviewees. Our discussions fueled my thoughts that led to this article! Here are the personalities I spoke with:

Julian Perry, CEO of Wine-Searcher – Katerina Axelsson, CEO of Tastry – Charles Slocum, CBO of Tastry – Pam Dillon, CEO of Preferabli – Andrew Sussman, CTO of Preferabli – Tristan Rousselle, Founder & Deputy CEO at Aryballe – Gérard Spatafora, Managing Director at E-Studi’OZ.wine – Vijay Bhagwandas, CEO at Tasting Intelligence – Magalie Dubois, Assistant professor at Burgundy School of Business. I also contacted 7 wine critics but none replied.

To be transparent: I am a lecturer in the MBA Wine & Spirits that Gérard manages in Bordeaux. Magalie and I share the same thesis director. Her thesis is about the “analysis of the recommendation on the wine purchase decision”. The idea of this article emerged while Magalie and I were discussing to prepare one of her conferences.

When I express my opinion, I use a more direct writing style, less "academic". I also changed the design of the paragraph with a preformatted writing and this grey background. It is easier to see the difference, and for you to skip the paragraph if you are less interested in my opinion.

Summary of the first article – Why are wine critics afraid of AI?: I have listed and analyzed four methods of artificial intelligence: 1. AI sommelier is not AI, 2. AI recommendations based on crowdsourced opinion, 3. AI recommendations based on individual preferences, 4. Sensory based AI – Person’s palate is as unique as a fingerprint.

Summary of the second article – Wine – the only industry where critics are so essential: but why? : You’ve already felt scared – unless you’re in the business – when faced with wine shelves in stores. This is normal. I analyze the marketing techniques and the reason why some experts work on this negative feeling.

AI and the French vineyards

To set the context, I will start this article with a personal experience I had a few months ago.

I was invited to speak at a conference with wine merchants from Bordeaux. Because of their history, the Bordelais are open to trade, especially with the British and the Americans. It remains nevertheless French, especially in this vineyard. The oldest families in the vineyard are nobles for whom tradition is the foundation that has allowed them to reach this level of wine quality. This is quite respectable.

So I had prepared my conference on “AI and wine distribution” with this in mind. I talked less about AI technology, much more about data and opportunities for Bordeaux negociants and chateaux. During the Q&A session, the discussion quickly turned to the power of platforms, especially American and Chinese, and influencers on the Internet. In short, everything but AI… It’s hard to get out of a state of mind where the only way to resist is to refuse such technology. It’s not easy to try to explain that instead of resisting, they could take the lead by getting rid of the hand of the platforms (web 2) to go towards the future (web 3) with blockchain and AI.

READ ALSO Blockchain: an Operating Platform for AI?

Is this really about tradition vs. tech?

This topic discussed during a conference at ‘Wine Paris’ looked like interesting. I started to watch the video. The subject was catchy! But here it is, the debate turned around the usual themes in France: traditional winemaking techniques, the rules regulating the  PDOs and PGIs, the development of organic / biodynamic / natural wines vs. table wines.

As you know, the French do not have an established trading culture. We, French, are a people of grumpy farmers, fortunately sometimes philosophical, especially around a glass of wine. Not very prospective or strategic in the medium term. Let’s say rather very centered on the production and then on the consumption.

Something tells me that it was not the same organization and panelists that were invited to the last Future Drinks Expo by the Beverage Trade Network. The topic was the same “current solutions and future plans for the beverage industry”. In San Francisco, the conversation was more about AI and blockchain. Ashwin Ram, Technical Director of AI – Google, was one of the speaker. You can listen his conference recorded at Future Drinks Expo 2022. To be fair, the “Act for Change” Symposium held in Bordeaux last June by Vinexposium seemed more interesting than the one it held during Wine Paris.

Back to the question: “Is tradition still the future of wine ?” Would the debate have been more interesting if it had opposed traditionalists or techies? I don’t think so. I am not sure that asking the question in this way is the right one. Indeed, it is not the one or the other. There is no reason to confront people, professionals, approaches of each one especially on these subjects: it would be quite sterile. Each professional is free to do as he/she wishes, according to his/her means and constraints.

I understand why a consumer would use a critic’s rating.  Now they can also use technology to see what their own personal rating would be, which is particularly powerful in making the right choice.

Pam Dillon, Cofounder & CEO at Preferabli

The collection and processing of data is involved in the whole process: on a small scale first in the production process, elaboration and tasting of wines. Let’s consider the implications for 3 groups of professionals: the winegrowers, the mass producers, and the prestige châteaux and estates.

AI to assist Growers in Winemaking

From the photographic analysis of the grapes, carried out before the harvest, we gather precious data (berry size, compactness, conformation and size of a bunch). All these elements give information on the health of the fruit. From there, we can envision a first scenario of the grape yield and the ripening profile of a parcel. In terms of logistics, this data also facilitates interventions in the vineyard and optimizes the quality of pressing.

AI will help small wine producers, for example from Greece or Bulgaria.

Gérard Spatafora, Managing Director at E-Studi’OZ.wine

Winegrowers will be able to apply AI in wine making, whether it is against rot (botrytis cinerea). AI also opens huge fields in the accompaniment of oenologists towards the sustainability (from one vintage to another) and the character of the wine. Campari, for example, has achieved consistent product quality (bourbon and vodka) using Rockwell Automation‘s solution.

Not to mention that winemaking is a sophisticated microbial process. By controlling biodiversity, this quantitative method would accompany the fermentation stage. Winemakers, the ultimate decision-makers, will help AI to give direction to the winemaking process, better controlled.

READ ALSO AI saved $10M worth of wine

With AI, winemakers will be able to anticipate possibilities of intervention on the tank. Moreover, if winemakers want to work on the aromatic structure of the wine, they will be able to retrieve a considerable amount of information with these pre-analysis tools. Winemakers will be able to make more informed decisions about the flavors, texture, tannins.. that they want to incorporate into a bottle.

We could predict the intensity notes, help the winemakers during the winemaking process.

Tristan Rousselle Founder & Deputy CEO at Aryballe
"Consumers need to regain confidence, not to be disappointed from one vintage to another, from one vineyard to another, from one appellation to another for the wines of the same winemaker. This trust, this relationship is essential.
The AI of the customer relationship (professionals and individuals) will allow the winemakers to delegate the order taking, the transport management, the answers to the consumers' questions. They will be able to focus on what they love most: working in the vineyard, producing their wine." - David BECK.

AI for Mass Market – The virtual sommelier: it’s over before it starts

The wine & spirits market is dominated by a few large groups (Diageo, Pernod-Ricard, Bacardi, Fortune Brands, Brown Forman, Constellation Brands…) that regularly compete for international brands. These groups have already taken the lead in AI, almost all of them use AI. Pernod-Ricard has been working on the implementation of AI since 2015 during its digital transformation plan. The French group uses AI to measure the effectiveness of its marketing campaigns (their communication budget is over 1 billion euros), for promotion and for the allocation of salespeople on the ground.

This emerging AI technology helps food and beverage companies manage their supply chain through logistics, predictive analytics and transparency. AI can also enable marketers and organizations to reach customers on a personal level, engage in deeper interactions, and improve their overall brand experience. AI can help analyze, monitor and infer customer behavior and sentiment.

Here is the list of AI applications for the food and beverage industry: food sorting, quality control and safety compliance, consumer engagement, production and packaging, maintenance, other applications.

The Rastal “Smartglass” is AI bases solution used in bars and restaurants. It is a drinking glass which is fitted with an NFC (near field communication) chip. Each NFC chip has its own unique ID number meaning that each glass has its very own identity.

In a smart bar, this kind of chip can communicate directly with a reader chip and the data can be transferred wirelessly into cloud. The platform analyzes the data and can process, for example, the type and number of drinks ordered as well as the times they were ordered in a clear and concise way.

"I think the 'virtual sommeliers' (a.k.a AI sommeliers) offered by retailers and restaurants will soon be replaced. There is still an entry cost to accessing AI technology for the public. These "AIs" know so little about you and have no way to really know you except to be permanently connected to your shopping list for single people (in a family, not easy to know who ate what, when and if it was enjoyed or not).
The AI embedded in your phones (Siri, Alexa...) are still not very present in everyday life. People are not yet ready for an optimal use of these AI in the form of virtual twin.  We should not scare the general public by launching things too quickly. The fact is that only your life companion - as the Samsung S4 advert to sell these smartphones said - could offer optimal personalization to your needs in all circumstances. Other AI-specific apps will have to plug in.
Large volume wine marketers (50% of American wines in the US are offered by 3 wine industry leaders) will perhaps be a concern for the development of AI for the general public. The transparency and learning assistance of sensory AIs could break part of their business model, if these AIs manage to emerge and pass the barriers to entry.
Will subscription-based box sales players be able to resist sensory AI? Of course not. When all consumers will be able to use personalization via their own sensory AI, all these business models will not exist anymore. These players will have to evolve quickly, as Palate Club did... start evolving their business model now." - David BECK.

AI for Prestige wines – Pricing is the key

Wine is a cultural product. Another aspect of wine is the speculative and financial side that is an integral part of this field, a quotation in the same way as works of art. In front of all these aspects, artificial intelligence will be able to support a tasting, to convert a customer by bringing the advice that a human could have brought under other conditions. Here is an overview of the latest advances in the field. AI techniques can offer an analysis of the price variation of great wines and predict the fluctuation over the next few months.

I believe that an AI tool could replace the critics for the primeurs. Besides, this tool would be adapted to the customer’s profile.

Gérard Spatafora, Managing Director at E-Studi’OZ.wine

Let’s take the example of La Place de Bordeaux:
Every year, between April and June, Bordeaux goes into effervescence. It is the “Primeurs” campaign. This is the time when the wines, still maturing in their oak barrels, are tasted by professionals, journalists and great experts. Everyone can form an opinion on the quality of the wines presented. The great experts give scores for each château, while the châteaux announce the en primeur prices of their wines. This price will allow merchants to immediately buy the wines and resell them to professionals and individuals.

The price of a wine is based on its reputation, its age, the quality of the vintage (as certified by the top experts) and the intrinsic quality of the wine (as seen by the top experts also in the form of scores) – Excerpts from : “At what price should you buy your wines? Ask the algorithm!“.

"The article talks about "fair price". On this prestigious and financial market for investors is it a question of fair price. Can an AI replace an expert and propose, in an objective way, a price for the primeurs. Intellectually, it would be ideal. This is even the most coherent model of using advanced AI (I really differ from the simplistic model of virtual sommeliers). This model works after wine critics have done their job.
But the reality of AI replacing the critics - tasting and rating the wines in their prime - is catching up with us. This market could be 100% predictive if there is no bias. I think there is. 
We have all heard about the methods used (few blind tastings, prepared samples...) or about the reasons mentioned by Lisa Perrotti-Brown MW, to leave Wine Advocate. 
Which chateaux would accept not to be in charge anymore, to be judged by a technology... sorry, I mean 'not to be in the human relationship anymore' when we talk about the amounts of money circulating between the actors of this prestigious market? Only the buyers will be able to demand that things change, that uncertainties become almost null (except for the transport of wines), that AI replaces les grands experts." - David BECK.

In the end it is always a question of generations

What is the real impact of new technologies on coming generations?
Born between 1995 and 2010, Gen Z represent 32% of the world’s population. They are more hooked on electronic devices and virtual spaces than any other group so far. For starters, there’s the issue of communication, which is not insignificant to many of us. Generation Z, while largely accustomed to digital communication tools, actually prefers face-to-face communication and the reverse is true for Millenials who prefer digital platforms for communication.

One of the most overlooked issues today, for example, is how Gen Z is trying to avoid the economic pitfalls that Millenials have become mired in. Generation Z is more invested in using new technologies to check out the best economic options, from researching the best prices for products before purchase to their adoption of technology in more aspects of daily life than previous generations.

While Gen Z’s constant immersion in technology that most of us simply didn’t grow up with may seem like a boon in terms of adapting to current cultural and social changes, the reality is almost counter-intuitive: Gen Z approaches new technology as an “extension of itself” rather than an addiction or compulsion.

The younger the generation, the more they expect to be engaged on an individual level. Sommelier recommendations lack scalability, and crowd-sourced data lack individual relevance.

Charles Slocum, CBO at Tastry

In a world where everyone has a platform on social networks, many Gen Zers want to stand out from the crowd and feel unique,” says strategy consulting firm OC & C, which conducted a large study of 15,500 participants in nine countries.
“Nearly a fifth of Gen Zers surveyed prefer to spend on experiences rather than products. […] This thirst for experience seems to be partly correlated with less materialism and concern for sustainability,” the OC & C study said.

"A mixture of sadness and anger, frustration is inherent to human nature. Nevertheless, AI will soon remove this negative emotion for the buyer. The sadness and anger when you have decided to buy a wine that does not suit you. Marketers use this emotion. Soon, it will no longer exist for the consumer.
It's a bit like learning. Nowadays, we are more inclined to protect ourselves against information overload and its corollaries (communication deception, marketing nudge, fake news, available brain time...). While my generation and the previous generations, information, knowledge is a luxury. It was necessary to search, to find, to make mistakes to learn.
The digital twin of the new generations will be able to predict, help to learn by avoiding the bias of the information overload. The different forms of AI that we have analyzed in this article - not to mention the new ones that will appear - will totally disrupt our society, our consumption and therefore the current consumerism ecosystem. The king is dead. Long live the king!" - David BECK.