Ministry of Finance Cloud Invoice · User research
User research journey
Competitor analysis, multisegment interviews, and accessibility—one thread from evidence to shipped patterns (Feb–Nov 2025).
User research · Competitor analysis
MoF app vs. mature third parties and LINE—where trust, habit, and speed diverge
Ministry of Finance Cloud Invoice redesign (Feb–Nov 2025): App Store signals plus screen-level benchmarks before IA and UI.
Landscape
MoF Cloud Invoice
Official2.5 · ~9K reviews
Trust and policy-complete features; weak on speed, spending insight, and approachable density for silver users.
Invoice Passbook
Third-party4.8 · ~320K reviews
Habit-forming UX—charts, widgets, draw alerts with sound; advanced corners can feel crowded.
Cloud Invoice (3rd party)
Third-party4.8 · ~240K reviews
Clean UI and gentle onboarding; strong on analytics and quick barcode access via widgets.
LINE invoice
LINE mini-appNo standalone store listing
Zero extra install inside LINE; simple flows and social context—advanced needs still spill to other tools.
Compared
- Official MoF listing vs. two high-volume third-party passbooks.
- LINE invoice mini app (no separate install, social context).
- Core tasks: first open, scan, carrier / barcode, prizes, spend views.
Patterns we mirrored (within policy)
- LINE-style onboarding: illustrated first-run steps, hints per job, hide/expand barcode widget to save space—mapped to ministry-first-launch guidance and clearer prize / donation paths + multi-language.
- Third-party passbook speed: multi-login (Google / Apple / biometrics), filters and swipe lists, lottery countdowns—mapped to in-app biometrics + password recovery and explicit prize deadlines / status.
- Passbook-class depth (Invoice Passbook tier): charts, quick barcode copy, widgets / wearables, optional sound for wins—mapped to optional home modules, fixed rails for scan + passbook, sensory cues only where regulations allow.
Cut or adapted
- No commercial partner slots or private campaigns in a MoF surface.
- Legal review on claims, promos, and data copy—not every “growth” lever.
- Backend limits: real-time spend depth and some integrations stay phased.
User research · Multisegment interviews
From public reviews to five audience groups—including low vision
Layered evidence for the Ministry of Finance Cloud Invoice App: what showed up in App Store feedback, then what diverged and converged across age, language, and ability in moderated research—with sketch personas to keep segments tangible for design decisions.
Layer 1 · Where the problem showed up first
2.8★ App Store reviews (Taiwan · ~Mar 2024–Mar 2025)
Findability
Prize redemption, carrier binding, and settings buried under labels that read like ministry jargon, not tasks.
Older adults
Small type, dense screens, and no obvious “next step”—families stepping in to finish flows on someone else's phone.
Review pull quote
“I know the feature exists—I saw it in a tutorial—but I can't find it again.”
Review pull quote
“My mom gives up. Too many words, too small. She only trusts paper.”
Layer 2 · Interviews across five audiences—including low vision
| Younger (18–30) | Middle-aged (31–50) | Silver (51+) | Foreign resident | Low vision | |
|---|---|---|---|---|---|
| First action on open | Scan — universal: paper receipt, QR on screen, or prize check—all roads led to scan first. | ||||
| Sharpest pain | Slow loads; hard to find the features you need | Forgot passwords / MoF verification codes | Don't know the next step; fear tapping wrong | Chinese labels & icons don't read at a glance | Can't complete core tasks without a caregiver |
| Info density preference | More on home — wins, charts, shortcuts | Balanced — signal without clutter | As little as possible — one obvious job per screen | Sparse — language-friendly hierarchy | N/A — predictability & VoiceOver clarity beat density |
Shared truth
Everyone converged on the same entry behavior: show the carrier barcode and scan paper invoices.
On the home screen, older and low-vision groups wanted fewer items, more whitespace, and vivid color for fast recognition—while younger groups wanted a denser dashboard—more modules, promos, and shortcuts—and a cooler, minimal palette. One layout; two incompatible defaults.
Yi-Ting · 18–30
Wants speed, aesthetics, dense shortcuts—barcode first, ads never.
Open full persona →
David · 31–50
Balanced power user—biometrics, reminders, reliable at checkout.
Open full persona →
Teacher Chen · low vision
Needs predictable layout, VoiceOver that doesn't shout, family assist today.
Open full persona →
Mei-Ling · 51+ · Silver
Maze-like flows, small type—needs one clear next step and trust when donating.
Open full persona →
Alex · Foreign resident
Dense Chinese UI and icons—needs calm hierarchy and obvious paths to scan and claim.
Open full persona →
E-invoice app · Synthesis
User pain by segment & cross-group findings
App Store reviews, interviews, and moderated tests—distilled into segment cards, a shared scan/barcode truth, and the home-screen density conflict that drove configurable modules in the case study.
Hard time finding features, couldn't find what I wanted right after open; Don't know the next step
App Store reviews and interviews pointed to the same pattern on open: people still had to hunt for what they needed. Prize redemption, carrier binding, and settings sat under labels that read like ministry jargon, not tasks—and the first screen felt sparse next to everyday consumer apps, so lottery status, spending, and shortcuts did not show up where younger users expected. Reviews also described small type, dense screens, and no obvious next step—families stepped in to finish flows on someone else's phone. In sessions, older participants froze at ambiguous controls: without a clear next step, they worried about triggering the wrong action on an official app—so they slowed down or passed the phone to family. The sharpest pain was findability and uncertainty about what to tap next—not lack of motivation to use the service.
“I know the feature exists—I saw it in a tutorial—but I can't find it again.”
On open, I still have to dig—I want the important stuff visible immediately, not a home that feels empty next to the apps I use every day.
Sharpest pain for silver users: uncertainty about what to tap next and anxiety about mistakes—not lack of motivation to use the service.
Forgot passwords / MoF verification codes
The account gate did its security job—but verification codes and password recovery became the wall people hit before lottery, carrier, or scan tasks.
Sharpest pain in moderated sessions: forgot passwords and MoF verification codes blocked access before core tasks—not lack of features.
Carrier binding and auto-transfer read as jargon without English or plain-language help
Without detailed explanations or an English interface, many foreign residents could not tell what carrier binding or automatic prize transfer settings were for—including that winnings can be deposited automatically. The labels assumed prior civic vocabulary, so the tasks sounded abstract instead of tied to daily shopping and payouts.
Participants who relied on English or mixed Chinese could follow the taps in a tutorial but still could not explain what carrier binding or auto-deposit did for them in real life—so they abandoned the settings or left prizes unclaimed.
Scan entry too small; no sound to confirm a scan landed
The original scan-invoice entry was easy to miss: even when people remembered where it lived, the control was too small to hit reliably. Scanning also gave no audio cue—so after aiming at a barcode, they could not tell whether the app had registered a successful scan.
Low-vision and blind participants often needed help to complete the same tasks younger users did alone—VoiceOver and visual hierarchy were not yet carrying the load.
Cross-group findings
Shared truth
Everyone has the same entry behavior of showing the carrier barcode and scanning paper invoices.
Conflict insights
On the home screen, older and low-vision groups wanted fewer items, more whitespace, and vivid color for fast recognition—while younger groups wanted a denser dashboard—more modules, promos, and shortcuts—and a cooler, minimal palette. One layout; two incompatible defaults.
User research · Accessibility
Accessibility Work – Screen Reader Focus
One of the most impactful parts of this project was working with visually impaired users. I interviewed two blind participants and recorded how one of them used the "before" version of the app with iOS VoiceOver. The video revealed key pain points—like unclear feedback sounds for focus states, buttons not being recognized as buttons, and sections being read as interactable when they weren't.
From there, I annotated the entire app: defining alt text for each component, setting correct reading orders, and ensuring VoiceOver communicated whether an element was actionable or purely informational. I also worked closely with engineers to ensure these annotations could be implemented. This accessibility work has become a major differentiator for the app. The revised version will be submitted to the official Accessibility Platform by Taiwan's Ministry of Digital Affairs in early 2026, with the goal of receiving national certification.
Short clip · Open on YouTube