Week One After the Mobile App Launch: What's Working, What's Already Changing

The JTX Markets mobile app listed on the Play Store on 30 August. Five days of real traders on real accounts gives us the earliest look at what the app is actually doing for people, which is almost always different from what we expected it to do during the build. This is a short, specific retrospective — what worked, what traders flagged, what we're shipping in response.

What the usage data says

Three numbers from the first five days that we didn't predict:

• Average session length: 11 minutes. Web mobile (jtxmarkets.com on a phone browser) averages 3 minutes per visit. The native app sessions are nearly four times longer. The main driver is the Activity tab — traders keep the app open watching their positions during active trades, which they didn't do on the web mobile view because the browser tab would lose focus every time they switched apps.
• FAB tap-through rate: 42%. The centre amber Trade button in the tab bar gets tapped on nearly half of all sessions. We expected it to be the second-most-used entry (after Home) and figured 20-25%. The actual number says traders think of the app as "trading first, browsing second", which is the right mental model but means the Markets tab is less important than we thought and the Trade FAB behaviour matters more.
• Mobile → web crossover: 68%. More than two thirds of mobile traders also opened the web dashboard in the same five days. The two surfaces are complementary, not replacement — people use the app for monitoring and quick actions, the web for deeper research and complex order setup. Both matter.

What traders flagged

Five specific pieces of feedback, each with the thing we're doing about it:

• "No push notifications when the app is backgrounded." This was the top request, by a wide margin. Shipping: a dedicated engine-events daemon that subscribes to private channels and routes fills, liquidations and position closes to APNs/FCM. ETA is end of October. (It did ship — the push notifications post covers the final rollout.)
• "Chart indicators are missing." On web you have SMA, EMA, Bollinger Bands, RSI and MACD. On mobile, the Phase 1 release shipped with no indicators at all because we wanted to validate the base chart in production before adding the overlay layer. Shipping: Phase 1 indicator parity on mobile within three weeks. (Shipped — see chart indicators are live.)
• "Can't edit a resting limit order." You could only cancel and re-submit, which loses time-priority and introduces a brief window where the order isn't live. Shipping: an Edit button next to Cancel that fires an atomic PATCH to the engine's modify endpoint. In the next release.
• "Position-add is one tap but doesn't let me specify size." The Add button on an open position fires a same-side market order at a default of 25% of current position size. Several traders wanted an explicit size input. Shipping: Add modal gets a size field with the same four quick-fill chips as the main order form. Next release.
• "Account switcher is buried." The pill to switch between Main and challenge accounts is on authenticated screens but not immediately visible. Shipping: moving the pill into the header on every screen so it's always one tap away. Next release.

What's working that we didn't expect

Two surprises on the positive side:

• The haptic confirmation on order submit. A small vibration on successful submit, a heavier double-tap on rejection. This was a last-minute addition that we almost cut. In the first week, three different traders mentioned specifically that the haptic made them more comfortable trading on the phone — the implicit "did the order actually submit" question goes away when you can feel the answer.
• Per-symbol timeframe memory. The chart remembering your preferred timeframe per symbol (BTC-PERP stays on 1h, EURUSD stays on 15m) is boring under the hood but turns out to be a trust-builder — nothing is more annoying than opening a chart you know well and having to re-select the timeframe every time. Several traders flagged that this makes the mobile chart feel more like a tool they own, less like a web page they're visiting.

What's deliberately not changing

Three features traders asked for that we're not shipping yet, with the reason:

• Dark/light mode toggle. The app is dark-only by design because trading screens are usually dark, and most traders have been looking at Bloomberg terminals and TradingView dark mode for years. A light mode is a maintenance tax (every component needs two visual variants) that we're not going to pay unless there's much broader demand.
• In-app order book depth beyond 8 levels. Space on a phone screen is the constraint. Showing 20 levels would make each level unreadable. The 8-level cap is a deliberate choice. A tablet-optimised view with deeper depth is on the roadmap but is a separate build, not a parameter tweak.
• Order flow / trade tape visualisation. This is a desktop feature that doesn't translate cleanly to a phone — the information density is too high. The mobile app will stay focused on monitoring and quick actions; deep analysis stays on web.

What this means for the next few weeks

The feedback-driven changes above are the near-term queue. Beyond that, two bigger pieces:

• TestFlight → App Store review. The iOS build is in TestFlight now. The App Store review process for crypto/perps apps is three to six weeks including the near-certain first-pass rejection cycle. We'll push through it as fast as review allows.
• Push notifications bridge. The engine-events daemon is the biggest pending feature, both because it was the most-requested and because it fundamentally changes how you can hold risk on the platform. Spec is done, build is in flight.

Access

If you have the app installed, pull to refresh in the Play Store listing to pick up the latest version as updates ship. If you're not on the app yet, the launch post has the install guide — Play Store direct for Android, TestFlight on request for iOS until App Store review completes.

If you've got feedback that's not covered here, we read every message. The product queue is actively being shaped by what traders tell us in the first weeks, and that window is still open.