25Live is a web-based room and event scheduling platform used by universities nationwide, including Penn State. Despite serving as the sole facility reservation system for a campus of over 42,000 undergraduate students, the platform is widely known for its steep learning curve and cumbersome booking workflow. In this research project, conducted in partnership with Ryan Ruiz, I examined the specific usability barriers that novice users encounter when completing core tasks on 25Live, and developed a set of evidence-based recommendations for the platform's developer, CollegeNET.
Role
UX Researcher
Client
25Live users across university campuses
Collaborator
Ryan Ruiz
Duration
Jan 2024 - May 2024
25Live is a critical piece of university infrastructure, meangint that without it, faculty, students, and student organizations cannot reserve rooms or schedule events on campus. Yet the platform consistently generates user frustration, particularly among first-time users who approach it with no prior training or institutional context.
The central research question driving this project was: Why do first-time users struggle to complete what should be a straightforward task, such as booking a room, and what specific interface and cognitive factors are responsible?
The project employed several complementary research methods, each targeting a different dimension of the usability problem.
Method 1: Learning Lab
Through 15 repeated trials with 5 novice college students, we measured task completion time to observe 25Live's learning curve — finding that while all participants improved significantly over trials and stabilized around 40–50 seconds after roughly 10 trials, initial completion times varied widely (120–400 seconds), revealing that the platform creates an inconsistent and steep entry experience for first-time users.
Method 3: Hierarchical Task Analysis & Cognitive Task Modeling
We decomposed 25Live's core workflows using HTA and modeled predicted completion times with KLM via Cogulator — finding that event creation requires a minimum of 11 steps and actual user times ran 3–5x over predictions, indicating cognitive overload as the primary performance bottleneck.
Method 2: Perceptual Interaction Lab
Using 5 Figma prototype variations incorporating principles such as Fitts's Law, Z-pattern, and pop-out effects, we timed participants locating the primary CTA button — finding that adding clear section headings alone improved discoverability as effectively as combining three design principles simultaneously.
Method 4: Google Lighthouse Accessibility Audit
Running Google Lighthouse on both the dashboard and event registration pages, we found that while accessibility scores exceeded 95, performance scores of 47–53 revealed significant front-end optimization issues contributing to slow load times that further degrade the user experience.
Read full reports here:
The primary usability failure in 25Live is cognitive overload, not visual noise or technical accessibility gaps. The platform presents users — particularly first-time users — with too many simultaneous decisions, too little semantic structure to guide navigation, and too few affordances to signal what the correct path forward is. The result is a task completion process that takes 3–5x longer than it theoretically should, even for users who are otherwise technically competent.
Based on the integrated findings, the following design interventions are recommended to CollegeNET:
Redesign the information architecture to respect working memory limits
The event creation flow should be restructured into a stepped or progressive disclosure format that surfaces only the information relevant to the current stage of the booking process, reducing the number of simultaneous decisions from 11+ to a cognitively manageable sequence.
Implement a robust heading and labeling system
The perceptual lab demonstrated that clear, semantically accurate headings have an outsized impact on navigation efficiency — comparable to or exceeding multi-principle visual redesigns. Replacing unlabeled content blocks with descriptive section headers is a high-return, low-risk improvement.
Improve system-to-mental-model mapping
Users should be able to find the HUB-Robeson Center by searching "HUB," not only by its formal facility name. The search functionality must be aligned with the language and mental models of the primary user population.
Redesign the primary call-to-action button
Applying Fitts's Law (larger target size) and a pop-out effect (high-contrast hue against the dominant color scheme) to the "Create an Event" button would meaningfully reduce initial search time and signal the button's primacy in the interface hierarchy.
Integrate contextual help directly into the interface
The current help documentation is accessible only through an external URL and a small, peripheral button. An embedded, context-sensitive help system — surfaced inline at the points where users are most likely to encounter friction — would reduce error rates and lower the learning curve for first-time users.
Prioritize front-end performance optimization
Reducing unused CSS and JavaScript, minimizing DOM size, and optimizing media assets would improve Lighthouse performance scores and reduce load times — addressing a systemic issue that compounds the interface's existing usability challenges.
This project was selected as the best class project of Spring 2024 in IST 331, and was recognized with the Fred Loomis Outreach Prize.
Following the semester, the course instructor, Dr. Frank Ritter, recruited me as a Learning Assistant for the course in the subsequent semester — a role based directly on the quality of the research and analytical work demonstrated in this project.