TED.com Index Page Redesign
Unearthing and improving the TED Talk discovery process for casual visitors to TED.com

Client
TED Conferences
‍
Duration
13 Months (Feb 2022 - Present)

My Role
Developed research plan, conducted interviews, created low and high fidelity wireframes.
‍
Tools
Figma

Project Summary

Challenge

Increase TED.com's retention by encouraging casual fans to return to the site.

Process
  • Generative user interviews
  • Prototyping index page redesigns
  • Analogous domains research
  • Usability testing index page prototype
  • High fidelity wireframes
  • Prioritizing new index page features
Solution(s)

Redesign TED.com/talks -- the 2nd most visited page on the website -- and other small tweaks to enable easier discovery of content.

Overview of the solution(s)

A redesign of the Talk Index page -- the catalog of all TED talks and the #2 most visited page on TED.com -- encouraging more serendipitous, less effortful discovery and exploration

Before

After

Process

explore

Discovery Research

Interviewed 5 casual TED fans to learn more about how they use TED.com and find videos.

draw

Wireframing

Explored design opportunities based on insights and pain-points uncovered in discovery research.

quiz

Usability Testing

Evaluated solutions for an index page redesign by conducting think-alouds with 3 users.

design_services

Hi-Fi Prototyping

Evaluated solutions for an index page redesign by conducting think-alouds with 3 users.

More about the challenge
TED.com hosts and publishes thousands of talks for people to enjoy. In 2022, one of their organization’s goals was to:
“Articulate why people come to TED’s site and mobile apps and focus on converting casual visitors into superfans. Increase O&O monthly retention to 10% in 2022.”

Among many other questions, our team asked, "What will encourage casual TED fans to return to TED.com?"

Casual Fans defined as:
- visiting TED.com < 1 per month
- No TED.com account
- May TED on social media
- Watches Talks on TED.com
‍

explore
Discovery Research

Reviewed existing research + documentation and orchestrated 5, 60 minute user-interviews via usertesting.com. Our guiding research questions consisted of:

  • Is there a “TED fan lifecycle?” Evolution of casual to superfan
  • What pain points exist on TED.com?
  • How do casual TED fans find content of interest on the internet
  • How and why do casual TED fans find and watch TED talks?

Research Strategy:

  • Review of web analytics of quantitative insights
  • Recruitment through usertesting.com
  • Development of testing protocol
Findings:
lightbulb

Users desire effortless exploration and serendipitous discovery of unexpected content of interest

“Sometimes I just want to see if there’s something interesting on the TED feed…I guess curiosity is the motivation…I don’t have any specific, like, ‘I want to…learn everything about something’ type of goal. I just want to enjoy the Talk.”
person

- S3

Supporting themes:

  • Discovery mostly occurs off-site
  • Users interested in casually learning something they wouldn't have search for intentionally
  • Exploration on site is cumbersome and frustrating
  • Time spent watching videos is on users' minds

‍

  • No desire for TED.com account
  • Users sees newsletter subs as a "trial"
  • 3/5 users mentioned they prefer listening to TED Talks
  • All users mentioned value of TED.com vs YouTube

To better convey these findings, I synthesized the data into a customer journey map depicting an approximation of the casual visitor's experience on TED.com,

Design Opportunities:

After presenting these insights with product stakeholders, I followed up with design opportunities that served to improve pain-points along the user's journey. These opportunities fell into 4 general categories, ultimately spawning projects of varying size across the 2022 and 2023 roadmaps:

Watch Next

Augment suggested content in watch next column with filters, topics, and more

Newsletters

Provide opportunities to share and customize email for curated talks

Topics Re-architecture

Reorganize topics to match user expectations and use as filters around TED.com

Index Page Redesign

Redesign page to reduce cognitive load when filtering for talks.

draw
Wireframing

While we worked on designs for other opportunities, the Index Page provided one of the most robust opportunities to improve user experience on TED.com.
‍
The index page is the #2 Visited webpage on TED.com (other than talk page) and was due for platform migration. Since it featured several qualitative pain-points for users browsing talks during our user testing, it appeared to be a great candidate for a redesign.
‍

The current state of the index page

lightbulb

"Topics" are most selected filter

lightbulb

~500k unique visitors per month visit the index page

lightbulb

~2% of all visitors to the index page navigate to page 2 of results.

From these insights, we experimented with different prototypes of the index page.

Responding to those questions, we created solutions inspired by analogous platforms such as Airbnb and Doordash -- other services that allow users to browse large catalogs based on a loose sense of what they might bein the mood for.

Our testing prototype

Proposed Solutions:

Improved Filtering

  • Floating Action Buttons
  • Re-architected topics
  • Supertopic launch pads
  • Look-ahead and smarter search

Improved Talk Presentation

  • Cards that emphasize useful metadata
  • Ribbon-based layout
  • Lazy-load

Suggested Browsing Paths

  • Trending content and topics
  • Thematic and seasonal content
  • Personalized recommendations
quiz
Usability Testing

I conducted 3, 60-minute think-alouds to test our prototype and determine which features assisted users in discovering new videos.

Our guiding research questions consisted of:

  • How to reduce effort when filtering to talks of interest?
  • How to reduce cognitive load when browsing and discovering talks?

Research Strategy

  1. Identify research questions
  2. Prepare prototype
  3. Recruit through usertesting.com
  4. Prepare protocol
Findings:
lightbulb

Users expect to explore content that is personally relevant with little exerted effort.

"So if they could be categorized in here rather than having a bunch of random topics, then that would make it easier…There’s just so many things happening here I don’t know what I’m looking and what could be interesting to me because it’s just a mix and match of everything.”
person

- G1

Supporting themes:

  • Filter Modal / FAB Invalidated
  • Topics Launchpads Validated
  • Search Reigns Supreme Exploration Method
  • Sub-topic Architecture Confuses

‍

  • Personalization is Table Stakes
  • Audio Remains Popular
  • Language Label Misunderstood
  • Surprise Me Sparks Interest
design_services
Hi-Fi Prototyping

I used the findings from our usability test to further refine our wireframes and decide what solutions to launch by scheduling and driving wireframing sessions with reps from Design, Product, and Engineering. The following high fidelity wireframe was presented to senior leadership before breaking down features into manageable releases.

Proposed Solution & Features:
  1. Improved Search and Filters - Implement look ahead search that also can parse topics.
  2. Topic Launchpads - Organize sub-topics by super-topic and provide a quick filter at the top of the index page.
  3. Trending - Popular content and topics, especially from other platforms, should serve as suggestions for users.
  4. Surprise Me! - Borrowed from our successful mobile app, this feature generates a random TED Talk.
  5. New Layout with Updated Content - Refresh of the talk page visual design, with the added functionality of updating content after each refresh to always provide something new and different to explore.
  6. Personalized Recommendations - Topics, Talks, and Playlists should always be relevant to a user’s watch history, likes, and other accessible data they have shared.
  7. New Talk Cards with Metadata - Talk cards should de-emphasize the thumbnail image and highlight other useful metadata to make it easier for users to find and scan content of interest
  8. Thematic and Seasonal Content - Playlists and content relevant to calendar year events can be suggested to users in easily scannable manners
  9. Lazy Load - Replaced pagination (2% of users clicked “page 2”) to more readily suggest more content.
Next Steps & Reflection

Not all of the proposed features could be included in the initial release. I worked with the Web Product Manager to prioritize and organize features into a scheduled roadmap in order to accommodate engineering capabilities and resources.

The project MVP incorporated a few changes as a result, including:

  • Changing launchpads from parent topic categories to the top searched topics
  • Maintaining grid based layout vs. ribbons
  • Swapping Lazy Load for Show More for accessibility concerns
  • Swapping Dark Mode for Light Mode

Negotiating with Engineering was a tricky challenge throughout the prioritization phase of this project. While we achieved an MVP and plan for subsequent releases, in the future, including engineering in early stages of prototyping could have streamlined feature deliberation.

Moving forward, our team looks to assess the new designs and adjust our future releases accordingly. Metrics we are considering as signals of success include:

  • Scroll depth
  • Talk card impressions (vs. clicks)
  • Abandonment rate
  • Session duration
  • Browse duration (time from index page to talk start)
  • Applied filters