Eye-tracking research that turned a clunky mobile experience into a clear path to global literature.
Words Without Borders (WWB) is a digital magazine that amplifies global literary voices, publishing translated works from diverse authors and cultures. Its mobile site is the gateway for readers to discover international literature, both renowned and emerging.
Ran eye-tracking sessions, synthesized behavioral and analytics data, and turned the findings into a research-backed mobile redesign.
The redesign cut bounce rate, sped up navigation, and surfaced multilingual content users couldn't find before.
206.7K of WWB's 417.2K users visited on mobile, yet every key engagement metric trailed the combined-device baseline. The desktop site felt finished. The mobile site felt unfinished. That gap is what this study set out to close.
…leading to high bounce rates and missed content across the mobile experience.
Eye-tracking showed scattered gaze, no dominant entry point.
60% bounce rate. Only 13.8% reached the bottom.
Only 37% could locate it. Most misread the tag.
Links looked like text. Just 25% knew what was tappable.
A research-driven approach across navigation and search, content visibility, multilingual discoverability, and the article-level experience where exploration actually happens.
Clear sections, trimmed metadata, 3–4 cards per screen.
One unified filter. Predictive search. Learn as its own tab.
Explicit language labels — French, Chinese — on every card.
Underlined links, spaced credits, sort-by filters on genre pages.
Aligning on goals, defining the questions, and choosing a mixed-method approach to study mobile behavior.
I met with WWB's Digital Director and editorial team to align on research goals, content priorities, and success metrics for the mobile experience.
Make it effortless to find specific content on a mobile-first browse.
Surface translated work and editorial features that were buried in the existing IA.
Tighten the homepage so readers stay long enough to discover something worth reading.
Three focused goals and the questions that drove every session.
Where are users focusing most on the homepage, and what's pulling their attention away from the navigation?
Are users confused by the “Multilingual” and “Learn” tags? Do they expect filtering before or after search?
What causes users to exit early, and which content moments would have kept them scrolling?
A mixed-method approach combining behavioral data with direct user feedback.
Performance baselines — bounce rate, scroll depth, session duration — plus heatmaps and scroll maps to visualize attention and drop-off.
Moderated eye-tracking with task-based exploration. Think-aloud protocol captured live frustration. SUS + open-ended feedback closed each session.
Mobile eye-tracking sessions in the Pratt Usability Lab. Real users, real gaze patterns, real friction.
Recruited a mix of new and returning users who reflect WWB's mobile-first, global audience.
First-time visitors recruited via the WWB newsletter, social channels, and the Pratt network. Age 18–32.
Graduate students and literature professionals already familiar with the existing site.
Moderated mobile sessions at the Pratt Usability Lab using Tobii Pro Lab. Pre-test (5–10 min), task execution across Homepage / Search / Multilingual (15–25 min), post-test SUS and open feedback (10–15 min).
Pratt Usability Lab
Eye-tracking with think-aloud
45–60 minutes
Tobii Pro Lab
SUS score, time-on-task, task success
Google Forms post-test feedback
The baseline scores set the bar — and surfaced exactly which behaviors needed to change.
A more granular look at the four tasks. Each focused on a specific feature of the site, so we could pinpoint exactly where the experience broke down.
Not everything was broken. Three things landed for participants and shaped what we kept in the redesign.
8 of 8 enjoyed the look. Style signalled credibility on first scan.
"This looks very professional, I'd think this is a reputable source."
6 of 8 liked the filter. The idea landed — the placement didn't.
"The filter style looks clean and clear, I like the option to filter by languages."
Every participant valued the no-paywall model.
"I appreciate not having a paywall, knowledge should not be priced."
Four patterns repeated across every session.
Only 2 of 8 scrolled to the bottom. Drop-off after 1–2 folds.
"There's too much going on. I don't know what to focus on first."
Only 3 of 8 found multilingual stories. Most read the tag as a translator.
"I thought 'Multilingual' would translate the article, not show different languages."
Filters weren't explained. Search lacked category guidance.
"I wasn't sure what the search was actually searching through."
Links looked like text. Eye-tracking caught hesitation and backtracking.
"I didn't realize the author's name was a link. It looked like plain text."
Cross-cutting the lab observations against GA4 and Hotjar to separate one-off frustration from systemic friction.
I pulled GA4 site-engagement data for the mobile audience and layered it against Hotjar scroll maps of the homepage. Two artifacts, one story: mobile readers drop early and bounce often.
Four insights showed up in both the eye-tracking sessions and the analytics. When a behavior surfaced in both sources, it stopped being a hunch and became a recommendation.
Gaze trailed off. Only 13.8% reached the bottom.
Misread in the lab. Near-zero clicks in Hotjar. Actively unused.
Browse by Topic drew the clicks; session time dropped at layer two.
Gaze hesitated on author names. Hotjar showed almost no clicks.
Five targeted design moves to cut friction, surface multilingual content, sharpen the article-level experience, and turn passive scrolling into intentional reading.
62% of users struggled to locate multilingual articles. They expected “Multilingual” to translate articles, and the tag had low visibility on cards and in filters.
Users interpreted the label as a translation feature. Tag placement was buried inside filters and missing from preview cards.
Replace “Multilingual” with labels like “Chinese / English.” Surface the tag on article preview cards. Swap “Multimedia” for “Audio” or “Video.”
75% of users were confused by the search experience. They expected filters before search input, not after, and mistook the “Learn” tab for a search-related feature.
No predictive suggestions, no category guidance, no clear separation between “search” and “learn.”
Predictive suggestions and trending searches on focus. Consolidate three filter menus into one. Move “Learn” out into its own nav tab for semantic clarity.
40% of users reported information overload on mobile. Large article cards, lengthy descriptions, and inconsistent alignment broke the scanning flow.
No section dividers. “Browse by Topic” showed one genre at a time. Mixed alignment disrupted vertical rhythm. Only 13.8% of users reached the end of the homepage.
Add section headers and dividers. Trim card metadata to title + genre + author. Show 3–4 cards per screen. Elevate “Browse by Topic” with all genres visible. Standardize left-alignment.
Only 2 of 8 users could identify clickable elements like author names. Links looked like plain text, and tightly spaced names led to accidental taps.
No underline, no contrast shift, no padding between adjacent names. Users missed navigation opportunities and tapped the wrong target.
Underline + add contrast on every clickable text element. Add padding between author and translator names to support touch accessibility.
Users struggled to explore long article lists inside a genre section. With no way to sort, the experience became overwhelm by default.
"It would be nice to get recommendations or top choices on the article pages since there are so many to explore, or even 'frequent flyers' of the WWB authors highlighted." (Participant 2)
Add a filter at the top of every genre page with options like Latest, Most Read, and curated picks, so readers can scan instead of scroll.
By simplifying navigation, optimizing content visibility, and clarifying search functionality, I created an experience that empowered readers to engage more meaningfully with WWB's diverse literary offerings.
3–4 cards per screen, Browse by Topic elevated, all sections clearly defined and left-aligned.
Three filter menus collapsed into one. Predictive suggestions on focus. Learn moved to its own nav tab.
Explicit language labels (French, Chinese, etc.) on every card. Multimedia split into Audio and Video.
Links underlined and spaced. Sort-by filters (Latest, Most Read) added to every genre page.
I look forward to delving further into the report and using it as a jumping-off point for improvements to the user experience.
Eric M. B. Becker · Digital Director & Senior Editor, Words Without Borders
Your work in this report is immensely helpful. Thank you so much for providing this thorough document. I can imagine WWB will be using the findings here for future improvements to their mobile offering.
Jessika Davis · Digital Humanities Fellow, Words Without Borders
Labels that feel obvious to designers can confuse real users, especially on a global platform like WWB. Watching readers navigate the site in real time was eye-opening, eye-tracking and think-aloud surfaced moments of hesitation that no analytics report could have shown.
What I learned: clear language leads to confident navigation, mixing data with live observation brings richer insights, and tiny tweaks like repositioning a tag or simplifying a filter can dramatically improve usability.
Good design isn't just about what looks clean, it's about what feels intuitive to real people.
