An evidence framework, not a chronological retrospective. Each transformation pairs a real product redesign with the constraints that drove it: regulatory disclosure, information density, or progressive complexity. The numbers are sourced; the trade-offs are named.
Institutional Visual Design
Trust, precision, and data density. Each principle below has a counterargument I've had to talk past in a real meeting — these aren't aspirations, they're positions.
Principle 01
Institutional users don't want delightful animations — they want pixel-perfect data accuracy. Every number timestamped, every chart sourced, every indicator verifiable. Sharp edges read as precision; rounded corners read as consumer.
Principle 02
Retail UX: "remove until simple." Institutional UX: "show everything, organised." Finlogix went from 200 points → 1,000+ per screen with 40% faster analysis time (n=15 traders, paired pre/post UI) — because hierarchy did the work density alone couldn't.
Principle 03
Every visual decision impacts perceived latency. Darker backgrounds reduce eye strain over 8-hour sessions. Monospace numerals improve scan-ability for tabular data. Triple-encoded direction (colour + arrow + sign) hits sub-200ms recognition — and stays readable for colour-blind users.
Principle 04
Legal requirements shape visual hierarchy. ASIC RG 268 mandates 12px minimum for risk warnings? I design around the constraint, not against it — high-contrast borders, strategic whitespace, pre-checked Legal approval. Zero design-related findings over 2+ years.
Principle 05
Every component must scale across 5+ products. Not bespoke layouts — reusable patterns. The ACY system's 150+ components share the same 8px grid, 6-colour semantic palette, and 4-tier typography scale. Result: faster implementation, 100% cross-product visual consistency.
Transformation 01 · ACY Securities
Homepage redesign for 100K+ global traders. Drag the divider to compare.
Trade-off · Brand vs. conversion. Marketing wanted more "personality" (illustrations, playful copy). I pushed back: financial trust requires restraint. We A/B tested "playful hero" vs. "institutional hero" — institutional won with measurably higher CTA engagement (internal A/B test, exact figures under NDA). In finance, less personality = more trust.
Transformation 02 · Finlogix
Information architecture redesign for active day traders.
User research finding — Michael Tan (intermediate day trader): "I used to spend 20 minutes every morning scanning for high-impact events. Now I see them instantly. Your colour-coding saved me 40% of my pre-market research time." The 40% figure is his self-reported estimate against personal hourly-rate calculation, not a portfolio metric.
Transformation 03 · TradingCup
Progressive disclosure for complex risk data.
The make-or-break moment · Legal collaboration. Initial design had 2 sentences for risk disclosure. Legal rejected it: "ASIC requires explicit warnings about strategy past performance." My compromise: kept the surface-level warning short (15 words), added an expandable "Full Risk Disclosure" link with Legal's required text. Result: 60% → 92% completion rate, zero Legal rejections post-launch.
System-Level Refinement
The transformations above are downstream of two upstream decisions: a 4-tier modular type scale and a 6-colour semantic palette. Both are documented; neither is bespoke.
System 01
Before: 7 different font sizes across products, mix of Arial / Helvetica / system fonts, arbitrary sizes (13 / 17 / 21 px). After: 4-tier modular scale at 1.33 ratio, Inter for UI / JetBrains Mono for numerals. Typography stops being a per-product decision.
System 02
Before: blue for everything (links, buttons, charts) — no semantic meaning, WCAG AA fail at 3.2:1. After: 6-colour semantic system — profit / loss / primary / warning / accent / neutral. WCAG AAA at 7.2:1 contrast on dark bg. Colour stops being decoration; it carries meaning.