Redesigning Taiwan’s e-Invoice app for accessibility & trust

One journey, one outcome

Challenge and context

Taiwan’s e-Invoice app is a nationwide public service: citizens store receipts, check lottery winnings, and manage rewards. When I joined Turn Cloud Technology Service Taiwan, the app was widely used but hard to use—people struggled to log in, didn’t trust the interface, and often gave up before completing basic tasks.

I led end-to-end UX for the authentication and core flows, working with government stakeholders and a cross-functional team of 10+ engineers.

One journey, three systemic barriers

Principles and decisions per gap

I treated this as one high-stakes citizen journey and translated research into three design principles: reduce cognitive load with step-based progress, increase trust with explicit system feedback, and improve discoverability with guided setup and clear IA.

Research and synthesis

I ran 30+ sessions—focus groups and 1:1 interviews—with teenagers, seniors, foreigners, and visually impaired users. I mapped where people hesitated, which terms confused them, and how they recovered from errors. The insight: many users weren’t “bad with technology”; they were confused by system language and flow structure.

System-level design principles

From research and stakeholder alignment, I defined a consistent language and hierarchy model: name things the way citizens do, reveal one next action at a time, make status explicit, and provide reassurance at critical moments.

Step-based progress clarityConfirmation and reassuranceGuided setup and discoverability

Flow architecture and iteration

I redesigned the flow architecture with explicit stage cues, rewrote feedback and helper messaging, and introduced guided setup patterns for reward banking and key actions. Prototypes were validated with users and used to align government leadership decisions.

Slide 1 of 3. Auto-advances; pauses on hover.

Page flow and navigation structure that simplified login and core tasks—shown across key architecture iterations.

How the product changed across the three gaps

Instead of showcasing isolated screens, this section maps shipped UI changes to each core UX gap so recruiters and stakeholders can see the logic from problem to solution.

Gap 01 · Cognitive load reduction

I introduced step-based flow, stronger hierarchy, and clear grouping so users always know what stage they are in and what to do next.

Step-based sequence in scan-to-claim

Hierarchy and grouping in ledger

Next actions are explicit

Clear progression in entry flows

Gap 02 · Trust and safety

I redesigned confirmation clarity, feedback messaging, and reassurance states to reduce fear of failure in invoice and money-related actions.

Secure entry choices

Clear status and confirmation

Transparent outcomes and value

Predictable structure builds confidence

Gap 03 · Discoverability and inclusive setup

For seniors and foreign users, I added guidance-oriented onboarding and setup flow patterns so critical features are findable and reward account setup is completable.

Findability of key functions

Guided entry and setup affordances

Inclusive information structure

Explicit next steps in rewards flow

Impact and validation

Results were measured not only by login success, but by clearer completion confidence, fewer support-heavy failure moments, and stronger alignment with public-service accessibility standards.

This project reinforced how much impact you can get from understanding diverse users, naming things clearly, and being intentional about the full entry-to-claim journey.

Validation before ship

I ran moderated prototype tests (one-on-one observation) with 12 participants covering diverse ages and contexts. Tasks spanned login and registration, carrier binding, invoice wallet and donation, scan and prize redemption, and spending analysis—mirroring the end-to-end citizen journey. Overall task success reached 88%, with concrete friction surfaced in verification and form entry that fed the next iteration.

For design leads & recruiters: this block is worth keeping—one image plus a short methodology blurb proves you validate before production and can tie work to task coverage and outcomes. Skip dumping the full Figma map; a single legible artifact (or this overview) is enough to signal rigor without overwhelming the narrative.