A widget system for a crypto exchange home screen

Freedx is a centralised crypto exchange, a new product in a category defined by Binance, Bybit and OKX.

The ask was Market Watch: a list of coins and prices on the mobile home screen. One widget, fixed position, same for everyone. I thought a single fixed list was the wrong answer for a product where different traders want different things, so I sketched a configurable version and took it to the CPO. He approved it.

Traders don’t want the same home screen. Someone holding only BTC wants a different first two seconds than someone tracking six alts and watching market sentiment. A fixed list serves the average user, and the average user doesn’t exist.

I looked at how the major exchanges handle the home screen, and at widget systems outside the category. The pattern is well established outside crypto, and less so inside it.

I’m also a trading platform user myself, which is where a lot of the judgment about what belongs on that screen came from. That’s not research and I don’t present it as research. It’s domain knowledge, and on a two-week timeline it was what I had.

Six widgets: Market Watch, Spotlight Coin, Hot Categories, Fear and Greed, VIP System, Crypto Loan.

The number wasn’t a design decision. It was set by what data engineering could actually serve, and by what could be built in the time available. There were more widgets worth having. We didn’t have the data behind them, so they didn’t get designed.

I’d rather state that plainly than present six as an editorial choice. The menu was bounded by the platform.

Most customisable home screens stop at on and off. This one has three layers.

Which widgets appear. The user adds and removes from the six.

Which view of that widget. Spotlight Coin can show one coin, two, or three. Market Watch has a dense view and a compact one. This layer is really about screen real estate. The user is deciding how much of the home screen a thing deserves, not just whether it’s there.

Which content inside it. Spotlight Coin lets the user pick which coins get highlighted. “Show me the market” means something different depending on what you hold.

Three widgets have variants: Spotlight Coin, Market Watch, Hot Categories. Three are fixed: Fear and Greed, VIP System, Crypto Loan. The fixed ones display one thing and there’s no meaningful second version of it.

One rule. A widget can’t appear twice in two different views. Two versions of Market Watch on one screen is duplication, not configuration, and it makes the screen worse. So selecting a view replaces, it doesn’t add.

Widgets sit in the order they were added, and users can swap positions afterwards.

The first version had a visible “+ Add Widget” button. I changed it to long-press, because that’s the convention users already know from their phone’s home screen, and a permanent button on a screen that’s supposed to be about the user’s own content is noise.

Long-press opens a menu with three options. Hide Widget removes the one you’re holding. Edit Home Page enters an edit mode where every widget carries a minus badge, for removing several at once. Edit Widget opens the view picker for widgets that have variants.

Two removal paths because they’re two different intents: fixing one thing, versus sitting down to rearrange the screen.

Hiding is not destructive. The widget goes back to the list and can be added again.

Long-press has no affordance. Nothing on the screen tells you it’s there.

So new users get a dismissible banner: hold widgets to customise your home page. That banner is a mitigation, and I want to be clear about what it doesn’t solve. Once it’s dismissed it doesn’t come back. A user who closes it on day one and then wants to change their screen in month three has to already know the gesture.

The trade was a clean home screen against permanent discoverability. I chose clean, and I’d make the same call again on this product, but a persistent entry point is the gap I’d close first.

The banner isn’t static. Once a user is in edit mode it switches to guidance for that mode, so the same slot teaches the gesture and then explains what to do with it.

A new user’s first home screen is empty apart from balance, actions, and a prompt to add widgets. They build their setup from nothing.

I’d change that. Shipping Market Watch by default would give a new user something to look at in their first session, and it would turn customisation into editing rather than building from scratch. The counter-argument is that an empty screen forces people to find the customisation, which matters when the way in is a long-press with no affordance.

It’s the open question on this project, and the first thing I’d want to put in front of real users.

The product calls this the P&L Widget Configuration, but P&L doesn’t behave like the six. It sits above the widget area, it’s always on screen, and it’s configured through a gear icon rather than through long-press. It’s part of the frame, not part of the system.

Three things can be turned on or off: today’s P&L, a small chart, and a large chart. The two charts are mutually exclusive, so enabling one disables the other. A user turning on the large chart shouldn’t have to work out why the small one looks wrong.

Engineering built it. It shipped to the app and is live, and I tested the build against the design.

I left Freedx shortly after launch, so I don’t have usage data to point to. What I can say is that it went from an unrequested idea to a working feature in two sprints, and that the system it’s built on holds six widgets, multiple views, and two themes without duplicating work.