Megan Reed

Product Design

Evolving wine pairing recommendations. Looking at how people are communicating, designing for that.

SommMe Chatbot
UX Strategy/User Research/Branding/UI Design

Discovering the Opportunity—

Let’s imagine an eager, amateur wine drinker. They have just started on their journey to understanding more about wine, something they have wanted to do for years. They have a full-time job, a busy family, a vibrant social calendar, so wine knowledge has always seemed like it would just add another to-do item to their list. They have never been able to feel confident about choosing a wine at a restaurant, let alone buying one for a nice meal at home.

They leave it to the waiter or salesperson to give the best recommendation and walk away barely remembering how to pronounce "Mer-lot".

Researching & Talking to Users—

We looked at how people are buying, consuming, and pairing wines and one of the major contributors was that consumers don't know where to start. Intimidating, daunting, and vast...the world of wine. One of the best ways to be successful in wine pairings is to better understand the food and pair to that.

Examining how users consume information about wines, we took results of interviews and surveys to drive insights to build the first flows.

Feedback sessions were facilitated to collate user feedback and insights into how the recommendations would function.

Hacking true insights

Our first version, we moved forward with a live chat feature. But, I found out that people are more interested in having an exploratory tool where they could access information that has been curated and personalized based on a customized criteria. As a result of user surveys and interviews, we used these insights to begin to design for a chatbot interaction.

Designing for Interaction—

The first visualization we created was the highest possible view we could offer, a more digestible form of a wine chart. This was the first step in moving away price, label, brand as a driving factor, and into an understandable view of wine pairing information.

With the relentless iterations between UX and UI, we took what we had learned through user stories and flows and produced wireframes that began to bring shape to the chatbot interaction.

I started looking at how people were obtaining information and communicating their preferences as inspiration for how we could make their data more approachable. In short order, the chatbot interaction came to life and this is where we focused our efforts as the UI began to take shape.

Branding the chatter

From the moment the user interacts with SommMe, it is important that the brand reflects a modern, approachable, conversational feel. The content, the tone of the chatbot, the logo, it is the whole of SommMe. One bad experience, across any screen, detracts from the brand, from the product, from the experience. The logo mark was designed to be understood as a stand alone representation of the brand, with communication and wine at the core.

Introducing SommMe—

Using chatbot interaction to facilitate accurate recommendations for food and wine pairings based on a specific sequence of questions and answers that results in a primary recommendation based on essential, international grape pairing connections - this is SommMe.

The goal is to inspire amateur wine consumers to learn about the vast world of wine through receiving approachable, expert recommendations to enhance the experience of drinking wine with a meal.

View Prototype on Invision

Bottom Line Insights—

Making wine pairing information understandable and approachable, is not only providing a useful tool for information, but it’s having a positive impact on the way that consumers enjoy wine.

As the next step, we’ll be able to expand our scope for what we can offer as useful knowledge to amateur wine consumers.

The more data that we have, the more opportunities we have to make that data useful. While we may expand our product scope, insights will layer on top of that data and be one of the core features of SommMe.