Web trading portal for
serious traders
Principal Product Designer
Design, Tech, Product, Analytics
Background
Only 5% of CapMint's users asked for a serious trading terminal on the web, but they placed 40% of all orders, so we built it, and it went on to account for 15% of orders on the platform.
I led the design direction end to end at CapMint, an F&O-focused trading platform. The result was a single-screen, fully customisable workspace where a trader can place, monitor and manage every trade without leaving the page.
The signal: a small voice with outsized weight
Only 5% of users were asking for a web terminal built for serious trading. On a feedback count alone, that is an easy request to park.
The volume told us a different story. These users were the platform's power users, and the orders they placed made up 40% of all orders on CapMint. Counting requests would have under-weighted them; weighing them by the activity they drove made them the most valuable group to design for.
The lesson I took from this: a minority signal is not a weak signal when the minority carries the platform's core behaviour.
A small group, a large share of the activity..
The business bet: make zero brokerage stick
CapMint was offering zero brokerage, and active traders wanted to use that window to trade as much as they could. For the business, that was a chance to turn high-frequency traders into long-term users of the platform.
A basic web view would not hold them. Traders who trade this heavily will move to whichever terminal lets them work fastest, so the portal had to be good enough that leaving felt like a downgrade.
Design principles we followed
I started by setting constraints for what the solution had to do. Four constraints came out of that:
One screen, every trade. A trader should never have to leave the workspace to see a chart, check positions, manage orders or scan the market.
The trader owns the layout. Instead of deciding the one right arrangement, we let each person build theirs.
Fluid over fixed. Moving, resizing and adding things should feel effortless, because serious traders rearrange their setup as the market changes.
Advanced, but calm. Dense, powerful tools on a neutral, uncluttered canvas so the data stays the focus.
Key design decisions
The portal is a canvas of widgets that traders arrange themselves. The default workspace brings together a chart, watchlists, scanners, positions and holdings, and orders, with total P&L and available funds always visible in the top bar.
A drag-and-drop widget canvas. Any widget can be dragged anywhere on the screen and the rest of the layout adjusts around it. Traders are not forced into a fixed grid, and nothing has to be manually resized to make room.
Customisation inside every widget. Each widget is itself a container. Traders can add tools or sections inside it, so a single panel can hold, for example, positions and holdings as tabs, or open and executed orders side by side. The Watchlist can carry multiple lists, and Scanners can switch between top gainers, top losers and gap-ups.
Built for multi-monitor setups. Many serious traders work across several desktops. Because widgets and tools can be arranged freely, they can spread their workspace across screens in the way that fits how they trade.
Named workspaces for different trading styles. The top bar holds switchable workspaces (Scalper, Turbo, Basket and Classic) with an option to add more, so a trader can move between setups without rebuilding them.
Action where the data is. Positions and orders carry their own controls: Exit, Reverse, SL/TP, Modify and Cancel, with Exit all and Cancel all at the top of each list. Trading decisions happen in the same place the numbers are shown.
Dark mode from day zero. We designed dark mode alongside the light theme from the very start, not as a later add-on. Traders spend long hours in front of screens, often across several monitors, so a comfortable theme mattered as much as any feature. Building both themes together meant every widget, chart and control had to work in each from the beginning, and colours that carry meaning (gains, losses, buy, sell, alerts) had to stay clear and consistent in both.
The complexity we solved
A terminal that is both powerful and fluid is hard to design, because every freedom given to the trader adds a way for the screen to break. These were the hardest problems.
Challenge
Why it was hard
How the design handled it
Free-form layout
Traders can drop a widget anywhere,
so every other widget has to reflow
without overlapping or losing usability.
Drag-and-drop on an adaptive canvas:
the rest of the layout adjusts
automatically around the moved widget.
Customisation without chaos
Letting people add any tool or section
inside any widget can leave them with
a screen they cannot read.
Every widget is a container with the
same simple pattern for adding,
tabbing and removing tools,
so the system stays predictable.
Everything on one screen
Charts, watchlists, scanners, positions,
holdings and orders all compete for
space, and traders must act on any of
them quickly.
Controls such as Exit, Reverse, SL/TP,
Modify and Cancel sit directly on
each row, with Exit all and Cancel
all at list level.
Different styles of trading
A scalper, a basket trader and a
classic investor need very
different setups.
Switchable workspaces
(Scalper, Turbo, Basket, Classic)
that can be extended with new ones.
Many monitors
Serious traders spread tools across
several desktops, so widgets cannot
assume one fixed viewport.
Widgets and tools are free to arrange
so the workspace can be laid out
across screens.
Dense data, calm interface
High information density usually
feels cluttered.
A neutral canvas, consistent type
and colour reserved for meaning
(gains, losses, actions).
CapMint
Easy Navigation
One click to switch tabs
Heavy customisation options
Fluidic layout
Sahi - Competetor
Coplicated Navigation
Multiple clicks to switch tabs
No customisation available
Rigid layout
Adoption was high. Once the web portal launched, it grew to account for 15% of all orders on the platform, from the same group of power users who had asked for it and were already driving a large share of trading.
The portal also compared well against competing terminals, which felt more rigid. Ours was more fluid and far more customisable.
Weigh signals by behaviour, not headcount
The 5% who asked drove 40% of orders. Who is asking matters as much
as how many.
Customisation is a design problem, not a feature list
Giving traders control only works when the canvas stays fluid and the defaults are good enough to start from.
Align the design bet with the business model
Zero brokerage made heavy traders eager to use the platform, so a terminal that held them there served both the users and the business.