
Solution and impact
- Fast-funnel onboarding that runs from sign-up to the first deposit, with wallet setup as a natural step, not a gate.
- A dashboard built around the balance and Deposit, with Scan QR, Send and Deposit one tap away.
- Plain language, a clear detail page for every payment, and recovery flows that reassure instead of alarm.
- A currency picker and RTL rules for Arabic, with the numeral format left in the user's hands.
2 audiences, one app:
Tether holders first, the unbanked next
Tether Payments Group is the payments arm of Tether. Its mission is to bring the billions of unbanked people into the dollar economy with payments that are safe, simple and cheap. The first users would be people who already hold Tether; the long-term audience is anyone, including people who have never used a crypto wallet.
- As few steps as possible. The brief set the bar at paying with a Visa or Mastercard in the US.
- No crypto talk. "Positioning us as a crypto company is going to do us a disservice."
- Volatile currencies in the first markets, and Arabic speakers whose habits differ by region.
- iOS in light and dark from day one, handed off to engineering with written rules.
In a financial app, trust is won or lost in the first session
People leave a wallet that feels like an obstacle before they ever fund it. So the core experience runs in one line from the first screen to money in the wallet: sign up, protect the account, back up the key, deposit. Each step asks for one thing and says why. When something goes wrong, a forgotten password or a lost phone, recovery is just as calm: 2 clear ways back, by email or with the private key.
Forgot the password
2 ways back are offered at once. By email, the link opens the mail app and a new password is set in 2 steps.
Lost the phone
The other way back is the private key saved at sign-up: paste it, import, and the wallet returns with its balance.
Paying should take a scan, an amount and one confirmation
"Open the app, scan a QR code… confirm your payment, and off you go." Scan QR stays one tap away on the home screen. The amount, the recipient and the fee sit on one confirmation screen, the code check comes only at the end, and every payment gets a clean detail page afterwards.
Mirroring is not RTL: I reengineered the architecture, not flipped it
Arabic speakers don't share one set of habits. Which numerals people read money in varies across countries, regions and even within one country: some users in the Gulf prefer Eastern Arabic numerals, ١٢٣٤, others Western ones, 1234. For an app this large, one mirrored layout would fail someone.
I synthesized the existing RTL rules into one Arabic version guide for the hand-off, so every screen solves the same cases the same way:
- Layout and navigation. The whole screen mirrors, including navigation bars, tab bars and sliders; Back flips, and transitions slide right to left. Icons that show direction, such as Send, flip too; carousels run the other way.
- Text direction. Alignment follows the language, not the mirror: an amount right-aligned in English stays right-aligned in Arabic, and Latin brand names don't break the right alignment.
- Numerals are the user's choice. People pick Eastern Arabic or Latin digits when they pick the language.
- Latin inside RTL. Email and password fields keep the RTL layout. Verification codes, the hardest case, render LTR within the RTL flow, aligned to the layout: filling the width or centered.
- Typography. Noto Sans Arabic, line height 10–40% above the English default, slightly negative letter spacing, and weights matched by eye across scripts.
Numeral format: Western vs. Eastern Arabic
Which numerals people expect varies by country, region and even within one country. So the choice comes with the language: Eastern Arabic or Western digits. Each screen has its own slider.
A variable color system that made dark theme a mode, not a redesign
Every component is documented for LTR and RTL, light and dark: a single source of truth for a dual-language, dual-theme product. Each screen below has its own slider: drag it to see the same screen in the other theme.
From 5 stakeholder interviews to a design brief the team could build against
I prepared the research: the interview plan with the Clay research team, and a competitor review of PayPal, Venmo, Nubank, Alipay, PayTM and Trust Wallet across onboarding, dashboard, deposit and withdrawal, and wallet security. After 5 30–40 minute interviews I synthesized the takeaways and took them straight into the designs, updating the screens as the picture got clearer.
| Takeaway | What it changed in the design |
|---|---|
| Users arrive mid-checkout and want to pay now | Fast-funnel onboarding: wallet setup reads as a natural step, not an obstacle |
| Funding the wallet is the biggest challenge | The dashboard puts the balance and Deposit first, with security and KYC as short next steps |
| "Open the app, scan a QR code… confirm your payment, and off you go" | Scan QR, Send and Deposit stay one tap away on the home screen |
| Distrust of crypto, low financial literacy | Plain language, a clear detail page for every dollar-backed payment, recovery flows that reassure |
| Language and currency vary by market | A currency picker, and the RTL rules for Arabic |
10% of Arabic-language users stayed because the app finally felt built for them
They could set the app up the way they read, numerals included, and every screen behaved the same way. The experience now serves 500M+ users.
What I take forward: when one audience has split habits, the answer is a choice and a rule, not a guess. Writing the rules down as a hand-off guide turns one designer's decisions into something the whole team and engineering can apply.




















































