epara — Trust-Centred Redesign of an Albanian Mobile Wallet LINK TO COMMERCIAL: https://www.youtube.com/watch?v=3QxLS-sTYJA LINK TO FIGMA PROTOTYPE: https://www.figma.com/proto/wl90u8s34Vzscmcl9pyaRd/epara-prototype?node-id=305-45&t=NIgBBYURmfiY8DXa-0&scaling=scale-down&content-scaling=fixed&page-id=0%3A1 Academic UX research & design case study Role: UX research, interaction design, redesign proposal Context: Master's thesis, Sapienza University of Rome (Design, Multimedia and Visual Communication) Platform: Mobile app (iOS) Tags: UX Research, UI Design, Fintech, Trust Design epara is BKT's mobile wallet, launched in 2023 as Albania's first bank-affiliated e-money product requiring only a national ID and phone number to register — no bank account needed. This case study investigates why a product backed by Albania's oldest, most trusted bank still faces adoption hesitation, and proposes a research-grounded redesign centred on trust rather than feature density. The Problem epara has a structural advantage most fintech products don't: BKT's institutional credibility, built over decades including surviving Albania's 1997 financial crisis. But the research found this asset is not being deployed by the product's design. Users encounter epara as a generic finance app first and a BKT product second, if at all, before adoption friction sets in. Research Methodology The study followed the Double Diamond process (Discover, Define, Develop, Deliver) and combined four methods: Observation. 5 participants in Tirana, spanning digitally-fluent young adults, working adults, and one participant with direct experience of the 1997 crisis, observed registering in the live app. Interviews. 9 semi-structured interviews (21-question survey), coded using Charmaz's two-step method, producing three user archetypes: the Routine Digital Payer, the Cautious Institutional Adopter, and the Loyalty-Anchored Traditional User. Heuristic evaluation. epara assessed against Nielsen's 10 usability heuristics, benchmarked against competitors EasyPay and POK. App review analysis. Qualitative coding of App Store/Google Play reviews for sentiment patterns. Limitations acknowledged in the thesis: all participants were Tirana-based, employed, and already held traditional bank accounts, so findings speak to hesitant/partial adopters among banked urban users, not Albania's unbanked or rural population, who are the actual target of the country's 2030 cashless goal. Key Findings Trust failures, not usability failures, registration steps (ID scan, OTP, PIN) were completed successfully by most participants. The real friction was interpretive: pauses and requests for reassurance before any data was entered. BKT is an invisible trust asset, users had strong prior trust in BKT but didn't connect it to epara until prompted. This was rated the most consequential finding. Fear of irreversibility, a recurring, crisis-linked concern that digital transactions can't be corrected once started. Language as an inclusion signal, fintech jargon made the product feel designed for someone else, especially for lower digital-literacy users. Heuristic evaluation flagged error recovery (Nielsen Heuristic 9) as the most severe violation (severity 4/4): error messages were diagnostic-only, with no corrective guidance or human escalation path. Design Response Redesign organized around Garrett's five planes of user experience, with three concrete outputs: Output 1. Trust-First Onboarding. Reorders the flow so BKT's institutional identity, regulatory status (Bank of Albania licence), and 90+ branch network appear before any data request — replacing a data-first sequence with a trust-first one. Output 2. Progressive Disclosure Home Screen. First-time users see only 3 core actions (transfer, top up, pay) instead of the full feature set on first launch, reducing cognitive load at the moment of highest disengagement risk. Output 3. Plain Language & Iconography Redesign. Rewrites every label and error message in plain Albanian, e.g. replacing a generic "Transaction failed. Error code 4412" with an explanatory, action-oriented message that names what happened and what to do next. A full high-fidelity prototype (20 screens) was built in Figma covering onboarding, identity verification, PIN creation, error states, dispute initiation, and referral, each screen tied back to a specific research finding (e.g. an "ID Pre-Explanation" screen was added specifically to answer the top hesitation point identified in observation: "who sees this?"). Outcome This is a research-led design proposal, not a shipped product change. The deliverables are: a tested prototype, a documented design rationale tied to named UX theory (McKnight et al.'s institution-based trust model, Fogg's Behavior Model, Nielsen's heuristics), and a set of findings, most notably that epara's biggest UX problem isn't a missing feature, but a hidden asset (BKT's own reputation).