Dynamic "Guests Also Viewed"
​Content driven by AI & ML

UX Research  -  Content Design  -  UX Writing  -  UX Design

Overview

At Disney, I led an effort to leverage AI and ML to dynamically serve optimal content
on our app and web pages. The project succeeded in:
  • decreasing detail-page abandonment by 7%
  • increasing conversion by 11%.
Picture
Situation

The Walt Disney World app and site feature hundreds
of single-topic detail pages for attractions, events, dining,
tours, etc.

We succeeded in making them easy to find, but we experienced
high site abandonment on these pages because we didn't
offer a compelling path forward to discover alternative activities.
Action

Content strategy leveraged UX research and page-level usage analysis to understand what
users were seeking on the pages and why they abandoned them.

The result: users abandoned detail pages when they were not offered a path forward.

Understand user
User Persona 

  • Identify users 
  • Align on primary persona
  • Create persona


Identify needs

Content strategy helped the cross-functional team refine the primary needs and strategy of the project.

Users needed a way to discover similar or alternative activities from single-topic detail pages. This was 
crucial when an activity was not available or in a different location from the user. 

User Need

User need
1) Discovery new experiences
2) Get personalized "things to do"
3) Easily book new activities 

Product Need 

Product goals/KPIs
1) Decrease abandonment
2) Optimize discovery
3) Increase conversions / revenue

Solution & Strategy

Content design strategy:
​1) Leverage AI for dynamic content
2) Optimize discovery of new content
3) Create new web/app component to showcase "Guests also viewed"




AI-driven content with predictive learning

What we did

We created a new "Guests also viewed" component at the bottom of every detail page/screen.
We used artificial intelligence to identify other attractions/activities with shared attributes (theme, location, experience, etc.)

How we did it

We applied AI and machine learning to serve alternative options to users who came to our detail pages but didn't find what they were looking for or could not book the experience.

  • We used machine learning to track and learn attractions users chose to view after leaving specific Detail Pages, then fed this data into an AI engine to "learn" attractions most relevant to other users
  • We used artificial intelligence to gather and align these choices to similar offerings on our sites and serve alternative content selections to guests in a new "Guests Also Viewed" section on the base of detail pages.

How we verified it

We iterated variations in layout, content and imagery to garner user feedback. 
Picture
Annotating solutions for app
Prototypes


Original detail pages

  • Our detail pages were dead ends -- if the user did not find what they were seeking, they began their search again
  • Pages limited discovery to just the one offering on the page
  • We failed to share the abundance of alternative options that were similar, nearby, related or popular substitute offerings
NEW AI-Driven Detail pages

  • Leveraging ML and AI, our new detail pages offered two new dynamic content sections: "Related Activities" and "Guests Also Viewed"
  • Data and analytics show users are continuing their discover by clicking these options to continue their journey and find alternative options
  • Users report an increased ability to find what they were looking for, based on options served by ML and AI
Picture
Results

The result
  1. Our detail pages were no longer "dead ends" because we now offered users a path forward to discover new compelling experiences.
  2. Content that was static became dynamic. Flat, non-specific content became personalized to deliver a more engaging experience. 

Metrics/data:
  • We increased post-detail-page discovery of attractions by 8%, measured by page visitations of 50 least-viewed pages.
  • We decreased dead ends on live detail pages by 25%, measured by usage of back button on detail pages over 6-month period.
  • We decreased page abandonment by 7.5%, measured on detail pages over 6-month period.
  • We received positive user feedback, measured by on-page feedback mechanism.

What I learned

This project taught me the role content strategy can play in strategizing both the content and the experience
on apps and sites. Our team was in the best position to understand the content we had, how users were
consuming it, and the tools to cross-link content to help users discover options they didn't know existed. 

Our users love the feature and report being able to "find more things appropriate for
our group" and "discover new things I never knew existed.

This was a win for the user and a win for content strategy.
Proudly powered by Weebly