A morning routine, redesigned.
A daily routine the spec didn't see.
Turkish users had a habit. Deposit Turkish lira into OKX in the morning, trade during the day, withdraw at night, move the funds to a local bank for overnight interest, then come back the next morning. The risk was never the withdrawal. It was the morning that didn't happen.
What the product did
Auto-swept idle TRY balances into a partner bank's savings product overnight. Returned them by morning. Same overnight interest, no manual movement, no missed mornings.
What I did
Led content across product flows, notifications, emails, and marketing. The work I'm most proud of isn't the copy. It's the shift in what we were trying to say.
The starting point looked fine. The market didn't agree.
The first value props read well in English. They explained the feature, hit the rate, and matched the global Earn playbook. I evaluated them against product goals and market context. On paper, they were reasonable.
I took them to the language managers and the local market team. The feedback was direct. Turkish users have lived through repeated currency crises and platform losses. They scan for proof, not polish, and assume marketing first. Phrases like "effortless earnings" read as a warning. Convenience matters only when it signals reliability.
I brought it back to the team. The launch language had two jobs at once: prove monetary value and prove trustworthiness. Trust had to be part of the product experience, not a tone overlay. We moved from explaining a feature to earning a habit shift.
"User mindset is cautious. High skepticism rooted in past economic instability and losses from online platforms."
Turkey language manager, research notes
The shift sounds simple. It wasn't. It meant rewriting onboarding, notifications, emails, and entry points so that the language stopped treating Turkish users like a translated audience and started treating them like the people they actually were.
Rules for a currency the style guide forgot.
The global style guide had rules for USD and EUR. None for TRY. I wrote a small system so product, marketing, legal, and localization teams could write about Turkish currency without relitigating the same decisions every sprint.
Full name in trust-building moments.
- Where
- Onboarding, marketing, first-time education.
- Why
- Full name reads as respectful and credible. New users need clarity, not crypto fluency.
ISO code in trading and portfolio surfaces.
- Where
- Portfolio screens, transaction history, balance lists.
- Why
- Matches global exchange convention. Traders scan it instantly.
Symbol in alerts and notifications.
- Where
- Push notifications, toasts, status banners.
- Why
- Scans in a glance. Saves space in tight UI.
Never combine formats on one screen.
- Where
- Design QA, the last check before any screen ships.
- Why
- Mixed formats break trust. Small doubt compounds.
How it travels. The same mapping works for USD, EUR, GBP, MXN. Decide the context tier once, then assign the format per surface. The next market launches inherited the framework instead of starting from scratch.
One notification, two relationships with the user.
The original copy wasn't wrong. It was written for the system, not the person reading it at 5 AM. This was the hardest notification to get right, because it lands on the worst-case morning: the user needs their money, and the bank hasn't returned it yet.
Assets under management at peak, in the first month. Around $1.29M USD.
The off-platform leak became an on-platform habit. Users who had been moving money out every night started staying. The morning routine that used to begin at a bank started on OKX.
The next market-specific launches (USDG Earn in the US, OKX Card in Brazil) started with the trust question first. Not "what does this product do," but "what does this market need to believe before they will move money." That shift was the actual product of this work.