Designing Options Backtesting From Scratch
A 0-to-1 UX build for web and mobile. Research, IA, and wireframes turned trader insight and competitive analysis into a full backtesting platform, before a single user existed.
Overview
Options trading is one of the fastest-growing but most intimidating corners of retail investing. Before risking real capital on a strategy, traders need to test it against historical data. Almost every existing tool for doing that comes with the same catch. Nearly every meaningful feature sits behind one all-or-nothing paywall.
PastStat was built to fix that. It was designed so traders pay only for the specific filters and predefined strategies they actually use, instead of gating the whole product behind a single subscription. It's backed by the largest, most specific library of predefined backtest plans in its category.
There was no existing product to redesign. PastStat was built from a blank page. My scope covered research, information architecture, and wireframes for the full web and mobile experience.
The Challenge
Zero-to-one meant no usage data and no legacy screens to react to. It didn't mean zero access to real traders, though. The stakeholders were option traders themselves, and conversations with their team added more perspective. A trader friend added further color on what risk means day to day, how much data and analytics matter to a trading decision, and what a correct, available strategy is worth in real terms.
What was missing wasn't domain knowledge. It was a live product and behavioral data to test decisions against. The category itself is inherently dense. Instrument, structure, timeframe, and risk profile all compound into real complexity. The risk was building something only power users could operate.
The tension that shaped everything downstream
Translate real but informal trader knowledge into a structured, approachable product. There was no live system to validate any of it against.
Research: Three Inputs, No Product
With no usage data to lean on, three things shaped the IA:
Stakeholders who were traders themselves, and conversations with their team
A trader friend's outside perspective on risk and what data means to a trading decision
A systematic audit of four direct competitors
That audit surfaced specifics that shaped decisions directly. Streak's entry and exit conditions were flexible, but complicated to set up. Its top-bar navigation also created real friction for mobile users. Option Alpha offered a strong curated strategy library and automated monitoring. But it gave traders no way to name their own strategies, and it was positioned as a tool for experienced traders, not beginners. AlgoTest's backtest visualization was a genuine strength. Its filters, though, were scattered across the interface, adding cognitive load instead of reducing it.
Here is the detailed competitor analysis:
Combined with direct trader input, this converged into a handful of product principles before a single screen was drawn:
Strategies need a name and short description to be understood at a glance
Filters are where users spend the most time. They need to be concise and intuitive, not a wall of fields
The results view is the payoff moment, and the clearest opportunity to out-do every competitor audited
Users need to save and personalize strategies for later
Risk mitigation matters enormously to this audience. Validating a strategy before committing real money isn't optional, it's the point
From Insights to IA
Those principles turned into the core flow: Sign up → Dashboard → Filter → Results list → Detailed view → Download / Save / Subscribe to paper trade. Every wireframe that followed built on top of this backbone.
Above: Userflow
Key Decisions
A progressive filter builder, refined over several iterations
The filter flow went through a few different structures before landing on categorized, collapsible groups: Primary filters, Option structure, Price action, Trend, Momentum, Volatility, Seasonal, and Events. That replaced a single long form. A conditional refine step sits on top for advanced users who want to combine conditions, similar in spirit to Streak's. It's a direct answer to the research. Filters are where traders spend most of their time, and AlgoTest's scattered-filter problem was the clearest example of what to avoid.
Initial explorations
Final design screens
Curated presets, or build your own
Beginners can pick from a library of predefined bullish, bearish, and neutral strategies. It's deliberately built out larger and more specific than any competitor audited. Experienced traders can ignore the presets entirely and build arbitrary multi-leg structures leg by leg: expiry, call or put, long or short, strike, ratio. Traders can pick one path or combine both. The flexibility was the point, rather than forcing everyone down the same route. It's a direct answer to Option Alpha's biggest limitation, a tool that only really worked if you already knew what you were doing.
The Detailed Backtest View: The flagship screen
Every competitor audited treated the results view as an afterthought: a table, maybe a chart. The research flagged this as the single biggest opportunity to differentiate. It became the flagship screen, with a PNL curve, a win/loss gauge, drawdown with recovery time, a full monthly and yearly returns heatmap, day-of-week returns, and a strategy-vs-benchmark comparison, all in one view. To make this screen beginner-friendly, I have added an info icon near complex metrics to give a quick glance at what they mean.
This is the payoff moment a trader arrives at after building a strategy. It's the place PastStat needed to feel obviously more capable than anything else in the category.
Detailed backtest view
A coin economy, not a paywall
Competitors in this space overwhelmingly use one model: a single subscription that unlocks everything, or nothing. The client's original brief followed that pattern. The design side proposed something different: pay-per-filter coins, layered underneath a separate subscription tier that unlocks whole features like Active Strategies. Every new account starts with a free coin allowance. That's enough to explore the system before any real spending decision is required, so the mechanic reads as a sandbox, not a tollbooth on a trader's very first action.
This was a genuine business-model negotiation, not just an execution detail. The reasoning was simple. A lower-commitment entry point would attract more first-time users than a subscription wall alone. It also directly answers the biggest gap the market scan turned up. Every competitor's best features sit behind one all-or-nothing wall.
Built around a sidebar, not a top bar
Streak's biggest usability complaint was structural: top-bar navigation that broke down for mobile users. PastStat's IA was built around a left sidebar from the outset. It kept a growing list of destinations organized and consistent across every screen: Dashboard, Backtest, Active strategies, Saved strategies, Paper trades, Subscription, and Settings.
Paper trading: Practice with dummy money first
One thing came up in nearly every trader conversation: the fear of being wrong with real money on the line. Paper trading was the direct product answer. It lets a trader run a strategy against simulated capital before committing actual funds. A saved paper trade surfaces in the same notification system as everything else, including live profit updates. The feedback loop closes the same way it would with a real position, minus the risk.
Designed for both platforms from the start
Every core flow was designed responsively for mobile alongside the web experience, not bolted on afterward. That includes onboarding, the dashboard, the filter builder, and the detailed view. The decisions are the same ones detailed above. What changes is density and layout for a smaller screen.
Tradeoffs
Charging coins per filter adds friction to a first-time user's very first action. In exchange, it trains more intentional, less scattershot strategy-building. To soften that friction, every new account starts with a free coin allowance, enough to explore the whole system once before any real spending decision is required.
Supporting both curated presets and a fully custom builder adds real interface complexity. In exchange, it gives traders genuine flexibility. They can pick a preset, build from scratch, or combine both, rather than being forced down one path.
Outcomes
Two to three months of research, IA, and wireframing culminated in a complete design system ready for build. It covers onboarding, the dashboard, the backtest engine, and subscription management, for both web and mobile.
My scope ended at handoff, so this case study focuses on the reasoning behind the decisions rather than launch metrics.
Reflection
Designing PastStat taught me that "no users yet" doesn't have to mean "no signal." The stakeholders were traders themselves. That proximity to real domain expertise did a lot of the work formal usability testing usually does. It just meant asking different questions, earlier, and trusting structured competitive research to close the rest of the gap.
The part I'd point to as the most senior work here isn't a screen. It's the coin-economy conversation. Taking a client's subscription-only brief and making the case for something more granular meant understanding their business goals well enough to argue with them productively, not just execute against a brief. I'd do that again before I'd default to just building what's asked.
