Design Lead
Design
Research
Testing
Through data-led, iterative design, we increased the total number of students landing on UAL course pages by 34%— while redesigning the search experience to be more usable, scalable and accessible.
Driving student recruitment is a priority for the UAL website, yet an analysis of the journey in 2023 revealed that 92% of users reaching the ‘Courses’ page as part of their discovery were abandoning, with only 8% reaching an actual course page; it is the second most visited page on the website, receiving over 3.4M visits per year.
Research
Working to understand the problem quickly - utilising quantitative and qualitative user research (from VWO funnel analysis to A/B testing, interviews and competitor analysis), the Digital Public Platforms UX team, in collaboration with colleagues from Student Marketing and Recruitment and in Short Courses teams discovered:
- The courses page was too long and complex.
- User flows to course pages were confusing and inconsistent.
- Search bars had poor visibility and accessibility.
- Filters were not scalable, leading to inconsistencies.
Design solution
Adopting an iterative, user-centred design approach, the team rapidly progressed towards:
- A simplified and reduced Courses Page.
- Streamlined user flows to create a more consistent experience .
- Improved search bar usability, responsiveness and accessibility.
- Redesigned filters with scalability in mind.
Results
After successfully releasing a beta to test improvements with real data and live users, Course Finder went live summer 2024 and has driven a dramatic 8%-> 64% increase in prospective students landing on individual course pages via the Courses journey.
+49%
Course page visits
+56%
Journey engagement
8% -> 64%
Task completion
Video Intro
Find out how I'm using Synthesia AI avatars to offer you a different way to explore case studies.
(2 minutes)
Intro
In early 2023, UAL’s Digital & Technology department was transitioning from IT-led initiatives to a more product and user-centred approach.
One top-level priority was building a Course Finder—an ideal chance for our newly formed UX team to show how user research, data, testing, and progressive iteration could drive business impact.
But this wasn’t just about a single feature; it was also a chance to tackle our broader search experience. We saw an opportunity to create search consistency across multiple platforms and break the cycle of siloed design within products.
The challenge? We had no product managers or senior leadership to guide us. So we forged ahead, determined to make sense of complex journeys and deliver a solution that met user and business needs.
Research
With limited resources (just me and a developer initially), we decided to let data and people guide the way.
Digital Probes and Quantitative Data
We set up a newly acquired data and A/B testing tool (VWO) and used it to create data captures to study user flows, funnels, and how people searched for courses. This gave us a baseline of what was actually happening with users.Competitor Analysis meets UserTesting
We enlisted a group of online testers and asked them to find courses on various versions of anonymised university websites (including ours). This approach showed us where we stood against other institutions reducing bias.Stakeholder Interviews
Finally, we spoke to stakeholders to understand why Course Finder had emerged as a project, what they felt was wrong with our current setup, and how we could align user needs with business objectives.
Challenges
High Abandonment Rates
Over 85% of users abandoned the journey after the first click on our Courses Page (think of the Courses Page as a main directory, guiding users into different course types and into individual course pages).
Course types were listed on an long and static page, burying high-value categories such as Short Courses at the bottom of the page (a key business vertical offering over 60% of all courses on our catalogue). As a result, only 17% of users ever reached the section.
Multiple branches and inconsistencies in both the journey and layout of similar pages left users confused, needing a major fix to our IA and user flows.
Issue #2: Inconsistent Search Components
The site had multiple search bars (Global, Course-specific, Showcase, and more), each designed independently.
Key searches were practically invisible in user tests: heatmaps showed minimal interaction, and many testers couldn’t even locate the main search bar until prompted. This also created important accessibility challenges.
Issue #3: Non-Scalable Filter Components
Filters were not mobile-friendly or scalable, and couldn’t support Short Courses (700+ courses). Short Courses ended up using a legacy filter system, contributing to the the search platform inconsistency.
We needed a filters that could scale and meet varying business requirements without splintering the user experience.
Design
Solution #1: Information architecture
Our existing Courses Page buried important sections of the business in long spaces and course offerings that were irrelevant to most users. We audited and cut out anything that was non-essential, resulting in a page with eight key course categories arranged by user needs and business relevance.
New course page architecture
We consolidated several branching paths into one centralised Course Finder journey.
Course finder as a way to unify a broken journeys into to individual course pages
A/B test results: Reducing page length boosted conversions to individual course pages by +42.69% supporting our hypothesis that the long page was causing users to drop-off.
Solution #2: Standardised search bar
Adopting Familiar Patterns
Instead of reinventing the wheel, we used recognized best practices (think Google) and pushed for a shift toward usability, scalability, and consistency over novelty. In an arts institution that values creative freedom and expression, this was a bold move.
A/B Testing for Buy-In
By pitting the old, stylized search bar against a simpler, more intuitive alternative, we saw a 284% initial jump in search bar interactions (measured in clicks), which later rose to 496% with a larger sample size.
New search bar and filters
Solution #3: Re-design filter components
Focus on Scalability
We collaborated closely with the developer team to ensure filter designs aligned with our search engine and metadata structure.
We ran workshops with student marketing and recruitment teams, to gain a better understanding of their needs and come up with ideas to build scalable filters.
Early Testing & Iteration
A new Junior UX Designer joined at this stage under my line of management, and together we ran rapid ideation and testing (from wireframes to interactive prototypes).
Some “cool” ideas flopped in testing; we learned quickly and simplified. Dynamic pills? Unnecessary for our us. Straightforward dropdowns with clear feedback? A hit.
Split URL Testing
We built a non-indexed test page to measure filter usage with real data. Results showed a 189% relative boost in filter interactions, rising from 3.21% to 9.28%, which helped us validate our new filters with quantitative data.
Outcomes
Shortly after launch, metrics soared beyond our highest hopes:
Task/ Journey Completion: 8% → 64%
Filter Queries: 34% → 51%
Search Queries: 5% → 15%
Individual course visits: 58% → 92%
Seeing a staggering 700% increase in the number of people completing the journey and 34% more visits on our individual course pages was a massive win for our tiny team. This win not only boosted user satisfaction and revenue opportunities but also trust on the UX team across the university. Our early-stage UX team improved a critical journey for prospective students while managing to demonstrate real value in a context of minimal leadership or guidance.
Key Learnings
💡 Design with Purpose
Grounding decisions in genuine user and business needs, robust research and testing led to tangible success—there was no guessing, when we launched, we knew it would make a difference.
💡 Communicate with Clarity
Aligning stakeholders through clear communications, transparent processes, and a straightforward data and user-led design rationale turned skepticism into collaboration.
💡 Flexible Frameworks Over Rigid Processes
We designed and tested each element differently depending on the needs, challenges and deadlines. It was by adopting frameworks like User-centred Design and Lean UX that we kept everything cohesive.
