As an ancient product, wine has a great legacy and tradition. And it is necessary to respect both if we want to provide a relevant solution for people. But we have also set out to push the limits of the category and explore new territories: Break the rules. Come up with disruptive things. Make the most of mobile by thinking about how it can serve each of us. The result is a unique and memorable conversational service, which is to help anyone progressively develop their interest and passion for wine.
What if buying wine wasn’t so complicated?
For less specialized consumers, buying wine can be complex. You have to understand the vintages, types of wood used in the barrels, nuances of flavours and smells, and know many grape varieties or the appellations of origin of each region, without forgetting the most important thing, how to pair it at a gastronomic level. Even for the most experienced, it is challenging to keep up with all the constant news and references.
There are already a series of apps focused on wine, but we found a design challenge: creating a more sophisticated proposal than the category offers. We want to surprise the users at every moment of the journey (approach, understanding, choice) and do it without forcing them to remember all the market references. To maintain coherency, consistency and interactivity at all service contact points with each customer, regardless of the channel.
Winebot arose as an assistant that simplifies the experience for everyone: from the most novice consumers to the most expert. Its mission is to help with unique suggestions designed individually for each person or occasion.
From reactive to proactive
One of the efforts that we try to build with this experience to achieve a medium-long-term relationship with consumers is anticipation: this is a stimulus to provoke interaction. Always be available to the client, learning at all times with each conversation and proactively offering answers. Using ANN (Artificial Neural Networks) to recognise, analyse, classify and predict responses.
Create a progressive connection
The connection begins by knowing the user’s expectations and learning from their habits and behaviours about their context beyond data processing. All aspects of the user that can be analysed are necessary to feed the learning engine.
Between the machine and the human
The relationship between the assistant and the client is based on interactions. They have to be convincing at all times, adapting to the user’s profile and knowledge of wine. It must balance the tone and wording so that they both are appropriate to the personalization that each moment of consumption requires.
What if you had an expert oenologist in your pocket?
The user can discover the best wines from the most renowned producers and wineries worldwide through mobile devices from different environments. It can be at home, in a restaurant, physical store, or in online commerce, thus eliminating barriers between the physical and digital world. Winebot is an assistant who helps you make decisions in crucial moments based on your level of knowledge and palate.
We want to generate an emotional connection and use different elements: depending on the year’s season, an exclusive selection, bestsellers for each occasion, type of humour or even type of atmosphere and music. We want to rethink new ways of understanding wine consumption beyond conventional formats or product types such as white, red, rosé or sparkling.
In a customer experience, we all expect advanced personalisation beyond sociology and demographics.
We have designed a journey based on contextualised data and moments of consumption. We have projected an experiential search through a conversational interface. We propose an NPL with the ability to accumulate a knowledge database based on the history of interactions with the user to continuously improve the service as if it were continuous training: remember each wine, recommending the perfect bottle whenever you need it.
Through this experience, we create reciprocal learning: Winebot provides a more personalised experience to the user, which is adjusted as the interactions grow. In the same way, the user obtains information more adapted to his tastes and interests, with which he can organically expand his wine knowledge.
A completely different purchasing model
Winebot is based on a new flexible subscription model. With access to a portfolio that ranges from the excellent references of award-winning wines to the most recent novelties. Active learning through AI with the user, based on their interactions, delivers recommendations based on personalisation-based criteria.
How does it know which wine to recommend? Thanks to the data we collect from the user, we can obtain an increasingly rich and detailed profile. Plus, we analyse the variants of the different aspects we need to take into account, from the context of the moment of consumption, the history of references, or the product or food with which to pair. We analyse this data correlation with the characteristics of the references in the database. In this way, the algorithm provides a series of unique wines.
Over time and with continuous learning, the assistant interprets and proposes to connect with the user’s expectations. Each recommendation is an exercise in proactivity to develop a steady growth in the user’s experience with the product.
The Future of Fitness
An overview of the emerging initiatives and user behaviours of the fitness world.