I ran mixed-methods research in collaboration with the business analyst and product manager — qualitative interviews with employees and customers, a quantitative survey to ARM’s entire customer base, and a structured competitor audit.

I designed a compact, flexible IA with five bottom-nav destinations — Home, Portfolio, Explore, Support, Profile. The structure could absorb any future product without restructuring.

I mapped a five-stage customer journey — Research, Onboarding, Purchasing, Post-Purchase, and Ongoing Engagement — documenting at each stage what customers were doing, thinking, feeling, and the design opportunities I’d target.

Customers used to log into separate apps for the entire ARM suite. I designed a single unified dashboard showing every product side-by-side — with allocation context, performance over time, and a recommendation layer driven by the customer’s risk profile.

The shift wasn’t visual — it was strategic. Once customers could see their whole financial picture in one place, cross-segment engagement became natural. This became the foundation for the +45% cross-segment engagement lift.

Usability testing surfaced a painful pattern: customers were only discovering platform tier limits at the moment they tried to invest more — drop-off at the worst possible moment. I redesigned the success state of onboarding to surface tier status upfront, with a one-tap path to upload Tier 3 documents.

I also split the dashboard for new vs. existing customers — the same screen had been trying to serve two very different jobs. Registration completion went from 52% to 76%.

Customers told me, again and again: “I don’t know what to buy.” So I designed a five-minute Investment Profiling assessment — five questions on income, time horizon, risk appetite, and goals — that outputs a personalised risk profile (Conservative, Balanced, Aggressive) and three matched product recommendations.

The Robo Advisor became the bridge between sign-up and first investment, and the foundation for the +40% retail participation lift across Securities trading and the self-directed product set. Re-takeable every six months as life changes.

Stocks customers are almost exclusively the Active Investor archetype — they need real-time data, depth on every position, and tools to manage what they already hold. A single page can’t do that work without becoming overwhelming. I designed a three-level progressive disclosure where each surface answers one specific question.

The Wallet became the financial hub. One balance, one set of controls, one place to top up from anywhere. I designed the funding sheet to sort payment methods dynamically by a 24-hour reliability score for that customer’s network (debit card, bank transfer, USSD, direct debit) — so the option most likely to succeed was always at the top. Payment success rate climbed from 32% to 47%.

Idle wallet balance earns its keep. Funded-but-not-yet-invested cash isn’t dead — it earns a base yield (up to 9% on USD) with a one-tap opt-in. The wallet itself becomes a soft money-market-lite product, creating a natural on-ramp from funded → earning → invested.

Auto-invest and Re-invest after maturity, surfaced on the product page. The team wanted auto-invest at sign-up. I held the recommendation back until the moment after a customer’s first successful manual top-up — when behaviour was established and trust was earned. Placing it on the product page (not the wallet) tied the commitment to the product the customer had just bought, so the recurring contribution carried the same intent as the original purchase.

Every product page in the launched suite (Mutual Funds, Stocks, Treasury Bills) was redesigned around one principle: four actions are always one tap away, the surrounding context adapts to who is viewing. View, Top up, Withdraw, Buy — the constant scaffold. Everything else shifts per archetype.

The First-Timer sees the product explained in plain language, with Strategic Asset Allocation and Fund Fact Sheet above the fold. The Active Investor sees balance, accrued earnings, performance, and transactional actions — education collapses. The Switcher sees a comparison strip (“Our Yield 21% p.a. vs Others 12–16% p.a.”) plus Disclosure and Fund Fact Sheet trust artefacts.

Archetype came from the Robo Advisor when a customer had taken it; from a small profile-signal model (age, holdings, activity, tier, time since sign-up) otherwise.

Only 8% of customers held more than one product. Leadership wanted a marketing campaign — but unification had already given us the better surface: the dashboard itself. I designed a contextual “next best product” card that led with the customer’s own money — a personalised LTV projection (“You’ve saved ₦340,000 this year; it could be ₦410,000 in 12 months”) instead of generic product marketing.

I A/B tested copy and placement; the soft, trust-respecting variant beat the aggressive one — cross-product adoption climbed from 8% to 23%.

After the first build I ran moderated usability tests — 10 participants over Google Meet, recording every session. The biggest finding: one dashboard served both new and existing users, so new users logged in to a $0 shell. I shipped, watched the data, re-interviewed first-week users, then greyboxed two directions with them before committing.

There was no design system at all. Five products, four teams — and every component (modals, status pills, money cards, fund rows) rebuilt from scratch, slightly differently, each time. No shared foundation, no governance, nothing to inherit.

I built it from zero — tokens, type, colour, accessibility — and grew the library component by component as the suite expanded. Just as important, I stood up the practices that never existed: weekly office hours, pairing with engineers on first implementations, every new pattern routed through me, and the why documented, not just the what.