Leading Finance Design at Gate

Finance products are hard to design for because one label can hide very different risks. In my first weeks I mapped the line into four areas: earn, borrowing, quant trading, and the rest, such as pay and rebates. For each product I wrote down who it is for: someone new to crypto, an intermediate investor, or an advanced one.

Then I used every product as a user on web, iOS and Android, and wrote down what I found.

1. Flexible savings had copy errors, extra information, and a different information structure on web and in the app.
2. Structured products had inconsistent colors and icons, and no obvious place to find trade history.
3. An older savings product had no responsive layout, and most of its content repeated the new Earn page.

I compared the headers of four exchanges, from the busiest to the simplest. Gate’s had eleven top-level options, six repeated entries, the same page under different names, and a pile of near-identical badges: a green flame, a red flame, low fee, zero fee, Hot, New.

The others each had something to learn from. OKX had cut its header down to eight items. Bybit’s trading flyout led straight to a trading pair. Bitget used a drop-down flyout for operations slots.

I wrote the direction down as four points: regroup related products, such as structured products with dual currency, and fold auto-invest into strategies; align the finance tab on web and in the app; use flyouts that open with the main action and leave room for operations slots; and decide whether trading pages and general pages need different headers.

The first question was how people find a product. One logic starts from the coin: if you hold USDT, show the best rate for USDT and let a tap open every product for it. The other ignores product names and sorts by risk: low risk, high yield sorted by rate, and structured products. We looked at how competitors, OKX in particular, handled each.

We chose using our own data and the competitor comparison, not by preference. The page that shipped uses both: coin recommendations at the top, and product-type tabs below.

In the proposal, three choices came straight from data:

1. Shortcuts. Which products get a shortcut at the top was decided by page views and unique visitors.
2. A smaller banner. I looked at how operations slots performed on the homepage and the savings pages, and shrank the banner to use the vertical space better.
3. Order. Products were ordered by business importance, performance, and how high the entry threshold is.

The 3.1 release moved the app to Flutter, so the spec could not leave anything to guess. The board below covers the home page; banners that swipe or tap; three kinds of tags, used only when needed; the three product tabs with their overflow and pre-start states; a loading skeleton; tab transitions; a tab bar that sticks to the top while the list scrolls; search; and dark mode.

Search opens with recommended coins and history, then narrows as you type. A coin expands to show the products you can buy for it.

Members of my team designed the Loan and Startup pages. I set the direction and reviewed the work.

For Loan, the page follows where the user is: has never borrowed, has borrowed before, or holds a position now. We simplified the position list, put the collateral ratio first, added a position detail page, and moved the actions people use most to the front, based on behavior data. For Startup, we cut the header down, enlarged the subscribe button, and removed noise from the cards so the main action stands out.

I wrote the finance team’s weekly review and a monthly comparison for every product line. After a release I compared like periods, and for each movement I wrote down whether the market, an operations campaign or the design most likely caused it.

Dual-currency subscription is a good example, although a team member designed it. Two weeks after the new flow shipped, conversion rose on both platforms. We named two kinds of reasons and kept them apart: the experience, which split the flow into steps and moved to the new Flutter build, and the product, which added coins to choose from and a reinvest feature.

The reports also say what we could not tell. When the 3.1 app launched, traffic in the app appeared to fall, partly because users who had not updated could not be counted yet, and tracking was still being added. I did not put that dip down to the design, and I do not claim a before-and-after result for the Earn page.

Besides the design work, I set up a few habits for the team.

1. A monthly competitor review. Every month the team shared one for earn, quant and copy trading, looking at three things: opportunities, experience improvements, and feature changes. The April review noticed, for example, one exchange expanding its first product row by default with USDT products open, and another changing its icon grid from scrolling to tap-to-expand.
2. A map of our partners. I mapped which product manager and engineer sat behind each product line, and drew up the second-quarter plan for each: earn, borrowing, copy trading, quant, Startup, pay, rebates and institutional VIP.
3. A file structure for the team. I reviewed how design files were kept, compared two ways to organize them, and settled on a compromise with one naming rule: platform or file type, requirement, then version or date.

Before I left I also worked on the app’s VIP fee page. It had to make the levels from VIP 0 to VIP 16 feel different without making the page heavy, and it had to show institutional users at the top levels that they were treated differently.

The direction was traditional and tactile: privilege, rank and status. Each band of levels got its own material: silver at the start, then gold, rose gold, blue, and black with a gold crown at the very top.

I drew the benefits two ways, as a table, which compares levels clearly but makes a long page, and as small cards, which use the screen better. For the moment of upgrading I worked on whether to send an inbox message or show a pop-up. The review feedback was that a big pop-up felt too heavy, and one note suggested showing it only on the VIP page.

I left before this shipped, so there are no results to report.

Four things about finance products stayed with me.

1. Group products by the risk a user takes. People do not know our product names, but they know whether they want something safe, something that pays more, or something structured. A page that follows that order is easier to read than one that follows our org chart.
2. Start from the coin the user holds. Someone with USDT in their account wants to know what USDT can earn, so the recommendation starts there.
3. Explain the hard products where the decision happens. Dual currency and structured products are the hardest to read, so the subscription page shows a payout forecast, and icons help explain the structured ones.
4. Read the numbers against the market. In crypto, traffic follows prices, so I judged a release by conversion and by comparing like periods.