UNDL Item Detail Page UX Update
UX Design
Instructor: Kat Chiluiza
Pratt Institute, 2026
Group Member: Shelly | Liwei | Sara | Kevin
In this case study, our team evaluated the item detail pages of the United Nations Digital Library for their usability to the general public. With a mix methods of eye-tracking, SUS score and post-task question, we are able to identify pain points and come up with 4 recommendation that help UNDL more accessable to general public.
Abstract
In this case study, our team evaluated the item detail pages of the United Nations Digital Library for their usability to the general public. Using data from existing stakeholder and website surveys, we identified four main research questions and conducted a mixed-methods usability study based on these questions, which included eye-tracking, SUS scores, and post-task interviews. Through Private Panel, we recruited eight academics with no prior experience using the United Nations Digital Library and conducted one-on-one eye-tracking tests with them in a controlled testing environment. Although users were able to complete most core tasks, their eye-tracking data revealed that unfamiliar terminology, inconsistent page layouts, and low text legibility caused uncertainty, reduced their confidence, and ultimately resulted in a SUS score of 46.4—far below the industry benchmark of 68. Based on these findings, we recommend improvements to labeling, context-sensitive tooltips, consistent disabled states, and typography that prioritizes accessibility.
When Finding is Not Understanding.
As a key entry point for users searching for records and publications, the United Nations Digital Library’s item detail pages account for 73% of total visits to the platform. However, users have voiced a series of complaints: “Who is the author?”, “What format should I check for this doc?” , and “I put it in my basket, does this get me my doc?” . Users seem able to find the files they want among the complex array of documents, but they lack sufficient confidence that they understand the page’s content even if they are at the right spot.
This brings us to the research challenge of this project:
How can we transform the United Nations Digital Library’s item details page into a webpage that non-expert users can more easily understand, navigate, and trust?
We further analyzed the data provided by Motomo and the feedback received from stakeholders, ultimately narrowing it down to four more specific questions:
In this project, I was primarily responsible for redesigning the upgraded user interface and led the pilot testing as well as subsequent routine testing. By combining various research methods, my team and I analyzed the quantitative and qualitative findings from the testing and ultimately designed a systematic plan to elevate the interactive experience, reducing friction between non-expert users and the United Nations Digital Library.
Seeing the Problem From Multiple Angles.
We use a combination of user testing methods to approach the problem from multiple angles. By analyzing existing data, we derive both quantitative and qualitative insights.
Eye tracking based on Tobii Lab.
People can be objective when it comes to telling the researcher what is going wrong, but their eyes movement tells a different story that adds evidence to support their claim. with eye tracking, we are able to capture how a user’s eye moving around the complex UI environment.
Using a protocol designed around the research questions, we conducted eight moderated usability tests. By analyzing real user data such as gaze paths and heatmaps, we identified 36 potential usability issues.
SUS (System Usability Scale).
Using an industry-standard post-test questionnaire, we asked users to subjectively rate the existing IDP in terms of efficiency, effectiveness, and satisfaction. The average score of 46.4 further corroborated the issues revealed by the eye-tracking analysis.
RTA (Retrospective Think Aloud) and post-task questions.
By revisiting the eye-tracking sessions with participants, we can gather additional quantitative evidence to shed light on issues we may have overlooked during our initial analysis.
The IDP Works on the Surface, but Breaks Down in Meaning.
The user testing results showed a polarized pattern: out of six independent tasks, two achieved a 100% success rate, while one had an 87.5% success rate and another a 75% success rate; the lowest success rate was only 62.5%. The significant variation in individual task performance highlights a major challenge currently facing the website: although most users were eventually able to complete the majority of tasks, certain specific sections proved far more difficult than others.
At the same time, the user’s SUS score has dropped to a low of 46.4, falling between “poor” and “unacceptable.” This score clearly highlights the challenges users face. The gap between task completion rates and users’ overall experience suggests that while users are often able to find answers, the effort required goes beyond what is necessary, and the user experience is neither smooth nor intuitive. Given that the industry average score is 68, a redesign of the user experience is essential.
After refinement, these were narrowed down to two main issues:
Turning Uncertainty Into Direction.
Supporting Evidence #1Five out of eight participants stated that while searching for the author’s name, they felt uncertain about whether UNCTAD was the author and would have preferred a real person’s name.
Insights #1The current website naming approach confuses users, who typically expect to see a person's name in the author field rather than an organization's name
Supporting Evidence #2When users try to find files to download, they often click on “format,” which actually refers to citation formats rather than file formats.
Insights #2
The terminology used on the website differs from common understanding. Users often get stuck for long periods or fail to find what they're looking for because they don't understand the meaning of certain terms.
Supporting Evidence #3Users get stuck on the previous page while searching for a written summary; even though they see the hyperlink to the correct page, they aren’t sure if it’s the right one.
“I looked at the details but I didn’t understand which one I could click to get the [record]. It was frustrating." - P6
Insights #3Although some users have already identified their goals, they still lack confidence in the results.
Supporting Evidence #4Even on pages without a download button, most users tend to look for a button or link labeled “download.”
Insights #4Due to missing files, some IDPs may not offer downloadable sections; this inconsistency in layout is obvious and confusing to users.
Supporting Evidence #5Users get stuck on the previous page while searching for a written summary; even though they see the hyperlink to the correct page, they aren’t sure if it’s the right one.
Insights #5Although some users have already identified their goals, they still lack confidence in the results.
Our Recommendations.
Recommendaion #1Relabel to words users are familiar with.Using more common terms rather than terminology specific to the UN system can reduce the time users spend trying to interpret the text, while also allowing them to navigate the content more quickly and confidently.
Recommendaion #2Provide tool tips to explain specialized terms.For terms that cannot be changed, we recommend placing tooltips next to them so users can quickly understand what each term means.
Recommendaion #3Consistent IDP layouts with disabled states and text.To reduce user disorientation, we recommend standardizing the IDP layout across record types and using disabled states to clearly communicate when specific actions are unavailable.
Recommendaion #4Standardizing sizing to meet accessibility needs.To improve legibility, we recommend increasing descriptive text to 14px and using regular font weight. Section headers should also be resized proportionally so the page remains readable without losing hierarchy.
More Accessible, Rather Than Obscured by Jargon.
Our research indicates that IDP does not have any fundamental issues. Users are generally able to complete tasks, but they must rely on intuition, repeated checking to understand the meaning of labels and links. Therefore, the key lies not in adding more features, but in making the existing system more intuitive for non-expert users.
By clarifying terminology, maintaining consistent page structures, improving feedback for unavailable actions, and increasing readability, the user experience no longer requires the user to know certain terminology to navigate through the website, they shall feel confident for each link they click into.
Designing Toward User Confidence.
Users shall feel comfortable recognizing key information such as author and identification number (in this case, the symbol).
Users shall feel in control even if some of the IDP is missing a downloadable file, understand the situation. Users shouldn’t be guessing why a file is missing
Users should be able to read the text easily to find the information they are looking for, along with the consistent hierarchy.
Our team synthesized various forms of evidence—SUS scores, task success rates, eye-tracking heatmaps, gaze maps, participant quotes, and session transcripts—to create a presentation that illustrates how users navigate the page and where they spend their time on the IDP
This handover gave the UNDL team a more complete understanding of the issue. Quantitative data revealed which aspects of the user experience were underperforming, while eye-tracking and qualitative feedback revealed the moments when users had to guess.
What I learned
This project taught me that a single testing method cannot fully identify a product's issues; only by combining various quantitative and qualitative research methods can we piece together the complete picture.
Although users completed the tasks during testing, the SUS revealed obstacles in the actual user experience. It was only by combining this with eye-tracking data, such as gaze maps, that we were able to truly understand where the friction points lay.
If this project were to continue, I would conduct more in-depth research into the differences between various IDPs and how users perceive these distinctions, to ensure that the United Nations Digital Library provides a smoother experience in future archives.