Designing chatbots
CASE study: Stubhub
Outcome
Growth strategy to increase audience and loyalty.
Omni-channel journeys to meet users where they are and use social networks to increase audience.
Background
StubHub is the world’s largest ticket marketplace with tickets available for over 10 million live sports, music and theatre events in more than 40 countries. The platform allows experience-seekers to buy and sell tickets through the website, mobile app and chatbot.
stubhub chatbot on facebook messenger
For building this chatbot our team partnered with the Facebook messenger team.
The UI was designed to be in compliance with Facebook design system .
The core conversation experience I patterned with our dev team, content strategist and product managers
Audience & Problem
Research shows that people largely use messaging platforms to plan activities with friends. The current experience of finding activities and co-ordinating with the group is fragmented across multiple platforms and channels of communication.
Approach
OBSERVATION
Within first 6 months, there were over 50,000 sessions on the StubHub chatbot platform, all achieved through organic traffic.
This showed us that people were engaging with the platform.
The most common conversation were people looking for tickets to specific shows.
How might we improve the planning and the discovery of events for event goers?
Event discovery in group chats
An integrated web experience that lives within the group chat.
We partnered with the teams at Skype and Facebook to design a web experience integrated within their group chat windows to allow people to easily share content without leaving the chat group.
Chatbot extension : user flows
An omni-channel event discovery experience that connects the website, mobile app and chatbot.
Chatbots on web for re-engaging buyers with high intent
Reduce Empty states
When a buyer came across a “No tickets” scenario on stubhub.com, we designed a flow that would send them a timely notification via our chatbot on Facebook as soon as more tickets become available
end-end omni-channel customer experience
Re-engaging buyers who come across empty state by offering a timely invitation to get notified via messenger.
What I learned about designing for chatbots?
Set expectations about what the chatbot can do.
People expect a higher level of personalization on chat.
Anthropomorphism is a common response towards chatbots.
Next Steps
One of the biggest insights for me during this process was that it was tough to get a group of friends to commit to an activity in time for buying event tickets.
I designed a proof of concept to show how a concierge bot could facilitate a convergent conversation to gather a group of friends towards a shared outcome.