AI to Individualize Pricing System (BtoB)

The wine trade has become globalized with the development of transportation technology. The use of artificial intelligence (AI) will significantly change the distribution channels, able to offer personalized prices by market, by actor of wine distribution.

The first article explains the major economic and societal challenges of artificial intelligence (AI) in the wine industry. We will then focus on the likely impact of AI on value transformation, more commonly known as disruption. David will then examine the contributions of AI on professional pricing in wine distribution. Finally, we will reflect on the impacts of wine price personalization for consumers at the dawn of AI-managed dynamic pricing.

Read also Price Personalization Through AI

New anti-competitive practices may emerge

Artificial intelligence (AI) offers a way to produce more reliable forecasts, recommendations or decisions at lower cost. It promises to generate productivity gains and help solve complex challenges. Wine producers’ pricing helps to explain how the diversity of their decision-making procedures is linked to economic and sociological relationships.
A 2015 study shows that machine learning methods can be applied to all Liv-ex 100 wines with an average predictive accuracy improvement of 15% over the most efficient traditional method.

AI can predict the highest price a buyer is willing to pay

The primary effect of artificial intelligence is to enable individualization of the online customer experience. The result is a much more organized sales process that ensures information accurately reflects prospects’ personal information, as well as product selection and pricing rules.
Advances in AI could impact business practices with the individualization of business rates. With AI technologies, companies can react according to the laws of supply and demand, profit requirement and externalities.
The price of wine and the profit margin depend on where it is sold. In the United States, for example, restaurants and bars have a profit margin of about 70% on wine, while retailers typically have a margin of 30-50%. Distributors and wholesalers, meanwhile, have a profit margin on wine of about 30%, and producers and wineries make approximately 50% gross margin.

Machine learning algorithms are used to predict the maximum price a player is willing to pay for a product. Prices are thus calculated according to each actor at the point of engagement. Natural language processing (NLP) allows campaigns to be adapted to linguistic and cultural contexts.

The OECD considers that “the practice of a firm using information voluntarily provided, gathered by observation or inferred about the behavior of individuals or their personal characteristics to differentiate prices between consumers on the basis of what the firm believes to be their willingness to pay”. Applied by sellers, this practice could lead some customers to pay less for a wine, while others would pay more than they would have if the price offered had been the same for all.

Artificial intelligence models could agree on prices

Companies may agree that algorithms create a collusive situation, as has been observed with the sale of posters on Amazon.
A second problematic configuration would be if two companies purchase a pricing model from the same vendor. The latter might then have an incentive to produce a model that does not threaten either buyer.
The third configuration would be the situation where firms use an algorithm, similar or different, that estimates that competition costs more than it brings in. The models could agree on prices without any such intent on the part of the people.

For his last column, David will talk about dynamic pricing managed by artificial intelligence in stores. Feel free to share with us in comments.