I’ve been working at Eventbrite as a UX designer for a 10-week internship last summer. I had the chance to work with incredibly qualified people and learn from an amazing design and product team. Here is one of the major project I worked on this summer.

Problem

Historically, Eventbrite has been focused on its organizers experience: ticket creation, sales management, events monitoring, etc. Today, as the company is getting more and more customer focused, Eventbrite is putting more efforts into making its events discovery process more effective.

Through several usability testing sessions on our current web and mobile platform, we found out that the way we categorize events isn’t working for our users. Today, we sort our events by category (i.e. “music”), subcategory (i.e. “pop”) and type (i.e. “festival”).

img1

Current categorization system on Eventbrite

Most of our users don’t understand the difference between a category and a type. This confustion is the result of an inaccurate taxonomy, and also a bad user experience on our current interface. When people open the categories dropdown list, they don’t see the “type” filter anymore, and they don’t see the subcategories at a glance (they are hidden under each category and revealed only if the category is selected). This is a problem for discovery because users are struggling to find specific events they are interested in. They don’t browse our events easily, and they don’t discover all the potential of our platform.

Goal

Our goal here was to create a better categorization system. We needed to improve the taxonomy but also the way we classify our events. Our new system should also live in an improved user interface. We embraced Eventbrite’s values which are to be customer-obsessed and mobile-first.

Process

We started our work with a full week of stakeholders interviews. We knew that many people at Eventbrite already worked on the subject and we wanted to gather all the work, thoughts, ideas around categorization that have been created within Eventbrite. In parallel, we studied all the previous usability testing sessions that have been realized on the subject. We also did a detailed competitive analysis.

stakeholder-analysis

We gathered all stakeholders’ ideas into a nice post-it board!

After this deep research phase, we started a design sprint. We carefully read the advice from Jake Knapp from GV who wrote the Sprint Book. We followed some of his precepts to be able to tackle our project. Our goal for this design sprint was to determine a new categorization system from an attendee perspective (focused on discovery) on mobile web (which was our first canal for discovery in term of traffic). With the help of members of the design and product teams who we invited in our design sprint, we ended up with a new categorization system which we designed, implemented in a new UI and tested onsite. All of this in only 5 days!

Here is the solution we came up with. We decided to kill subcategories and types, which were bringing too much confusion. We established a new set of high-level categories, we allowed attendees to select multiple of them, and most of all, we added a tagging system to make the search more specific. We also added attributes which describe the event in a more practical way (i.e. “Kids-friendly”, “Parking onsite”).

img2

Overview of my designs for discovery on web mobile and desktop

After the design sprint, we focused our work on adapting our solution on desktop. Then, we focused on the organizer side. How will organizers create tags ? We decided to think of a multi-level implementation for our tagging system. We’ll start with the use of a free-tagging system to build a trustful word bank.

img4

Overview of the tagging system from an organizer focus

During all the process of re-designing Eventbrite categorization system, we tested our solutions on users (remotely and on-site). After all this design phase, we challenged our system with the needs of other Eventbrite’s stakeholders such as the marketing team and finance team who uses the events categorization system internally.

Conclusion

At the end of the summer, we presented our solution to Laurent Sellier (VP of Product at Eventbrite), Brian Beaver (VP of Design at Eventbrite) and Eventbrite product managers. We received great feedback and we felt that our project will have a significant impact on Eventbrite products. Some of the steps of the implementation are already in the 2017 roadmap!

I worked on this project with a dream team composed of two other interns (Eric Nalin, UX research and Lucy Minott, product manager). We were all under the supervision of our managers: Jeff Zundel (Senior UX Architect), Kristine Holst (Senior UX Researcher) and Whitney Ryan (Senior Product Manager).