Case Study

PaoTang Homepage Revamp

What 14 users actually thought — and what the team didn't expect to hear.

A qualitative research study on the homepage of PaoTang, Thailand's leading government-linked mobile banking super-app. The headline finding wasn't a verdict on a layout — it was evidence that the team and its users didn't share the same mental model of what the app is for.

UX ResearchFintechArise by INFINITAS

On confidentiality. This project involved an unreleased design for a live product with tens of millions of users. At the request of the Arise DX team, the new homepage designs are withheld and specific redesign recommendations are kept out of this write-up. What follows focuses on my research process, contributions, and the shape of the findings. Happy to walk through the redacted detail under NDA.

Role: UX Research Intern — wrote the discussion guide, owned synthesis, moderated and note-took across sessions
Company: Arise by INFINITAS (Krungthai Bank), Bangkok
Product: PaoTang — Thailand's leading government-linked mobile banking super-app
Timeline: July 2026
Method: Qualitative in-depth interviews (online + on-site), 1-hour sessions
Participants: 14 across three groups — students, office workers, elderly (ages 21–64)
Deliverable: Task write-ups, prioritised recommendations, 30-slide stakeholder report

Research overview — qualitative study, 14 respondents across 3 groups, in-depth interviews, July 2026 — Research overview from the 30-slide stakeholder report

The setup

PaoTang is used by a large share of the Thai population — mostly for government programmes rather than everyday banking. The product team had a proposed homepage redesign and a reasonable question: is the new layout better than the current one?

My role across this study was to help answer that — and, as the research unfolded, to surface the more important question underneath it. The headline outcome wasn't a layout verdict. It was evidence that the team and its users didn't share the same mental model of what PaoTang actually is, or who it's for. That gap was the most valuable thing the research brought back to the room.

Brand perception — the identity gap was the strongest signal in the data

The Challenge

Evaluating a homepage isn't just an aesthetics question, especially for a product this large and this loaded with meaning for users. Three things made this study genuinely difficult:

Breadth of user. Students, office workers, and elderly users have very different habits and mental models. A homepage that works cleanly for one can quietly fail another.
Mental models, not menus. The real question was whether users understood how the homepage's information was organised — not just whether a button looked nice or a layout felt modern.
Aesthetics vs. clarity. A redesign can feel more modern while being harder to actually use. Separating "I like how this looks" from "I can complete this task" was central to getting useful data.

Sub group analysis — three user groups (Student, Office Worker, Elderly) with findings on promotion banners, service categories, and rewards widget visibility — Sub group analysis — how the same homepage read differently across three user groups

What I owned

Research tasks and moderation

I drafted an early version of the moderation script — the final version used by the team was refined from there. What I did design fully were the two card sorting tasks run across all 14 sessions. In the first, users sorted PaoTang features into must have, nice to have, and can remove. In the second, they were given a blank phone screen and a set of hi-fi buttons, icons, and features, and asked to place whatever they wanted on their homepage. Both tasks were designed to get past "I like this layout" to what users actually valued and how they thought the app should be organised.

I independently moderated one complete end-to-end session and provided live note-taking support across the rest — capturing SEQ scores after each task, behavioural observations, and verbatim quotes. Sitting through that many sessions, across three very different user groups, showed me how differently people approached the same screen, and how often what a user said diverged from what they actually did.

Synthesis and recommendations

I built the primary cross-participant rainbow sheet from scratch — a synthesis matrix mapping every participant against every insight, 42+ rows in total. It let me separate confirmed patterns from one-off reactions, and surface contradictions between what different user groups expected from the same feature. That distinction — between a quote and a finding — is what makes recommendations defensible rather than just compelling.

From the synthesis, I developed the prioritised recommendations and built the full 25-slide research report delivered to the team.

Summary verdict — 10 of 14 users preferred the new homepage overall, but information clarity scored lower than the old design. Recommendation: combine new visual direction with old design's clarity. — The research verdict — users preferred the new design's look, but the old design's clarity won on usability

Presenting to stakeholders

I presented the findings directly to a room of ten stakeholders — UX designers, business unit representatives, and product owners. That was the moment the research had to do its real work: not just be accurate, but be clear and persuasive enough to reframe how the team was thinking about the project.

The findings I brought to the room:

A target-user gap. The strongest repeated signal wasn't about layout — it was about identity. Users' perception of who PaoTang is for didn't match the team's assumptions. Before optimising a homepage, you have to define who it's optimising for.
Aesthetics were outpacing clarity. The more modern direction was appreciated visually but repeatedly traded away affordance and legibility. Keep the ambition; resolve the clarity costs.
Usage predicted value better than demographics. How heavily someone used the app shaped what they valued far more than age or job title — a useful corrective to the age-group framing we'd started with.
Unresolved mental models. Users struggled to understand how the app's multiple balances and account types related to each other. A restyle alone wouldn't fix that — it's a structural problem that needs a structural answer.

Key takeaways — four actionable conclusions: define a target user, balance aesthetics with clarity, prioritise wallet and rewards above the fold, expand interaction preferred — Key takeaways from the stakeholder presentation — delivered to 10 people across UXD, product, and business

Reflection

The finding I hadn't expected was the one worth leading with. We came in with a layout question. The brand perception data told a different story: users described PaoTang as orderly, bland, functional — a government service you use because you have to, not because it's yours. A product trying to serve everyone ended up resonating with no one in particular. That was the more important question underneath the brief, and it only became clear once the synthesis was done.

The usage finding was less of a surprise and more of a pattern that made sense once I saw it. We'd gone in thinking about three user groups. What the data actually showed was that how much someone used the app mattered more than who they were.