Design Systems

A multi-brand design system built to reduce UI drift, speed up brand delivery, and create reusable product infrastructure across a large ecosystem.

Design systemsTokensGovernanceScaleMulti-brand UI
Leadership · Scale · Systems · Governance
Role
Lead UX/UI and design systems owner responsible for shaping the system direction, token architecture, component strategy, governance model, and adoption approach across design and delivery teams.
Project type
Multi-brand design system platform
Featured outcome
60–65% → <20%UI inconsistency reduced

Proteus is a multi-brand design system platform designed to bring consistency, speed, and governance to a complex product ecosystem. Multiple brands shared the same product foundation, but visual and behavioural drift had made delivery slower, QA harder, and rebrands unnecessarily expensive. The work focused on building a token-first, component-led system that allowed brands to express themselves while keeping the underlying product experience stable and reusable.

Proteus — hero showcase
Problem

The challenge

Design and delivery inconsistency had become a scaling problem. Multiple brands shared one product core, but each brand had drifted visually and structurally. Spacing, typography, components, states, and behaviours were inconsistent. Rebrands often became rebuilds, and quality depended too much on who last touched the UI.

Proteus — problem visual
Process

How I approached it

The process started by identifying where inconsistency appeared most often and why teams were bypassing shared patterns. From there, the system moved toward a token-first model. Semantic tokens defined product intent, while brand tokens handled visual expression. Components consumed the semantic layer, allowing the same component to support different brands without being duplicated.

01

Audited repeated UI patterns and inconsistencies

02

Mapped common component and layout needs

03

Defined a semantic token architecture

04

Connected brand expression to semantic intent

05

Created reusable components with states and variants

06

Introduced governance for intake, review, release, and QA

07

Aligned design and engineering around reusable rules

Proteus — process visual
Proteus — key decision visual
Challenges

Trade-offs

The main challenge was balancing flexibility with control. Each brand needed to feel distinct, but the product still needed to behave consistently. Too much freedom would recreate the original problem. Too much rigidity would push teams to work around the system.

Proteus — solution detail
Solution

Final direction

The final solution was a token-first, component-led design system. The foundation separated product intent from brand expression. Components had clear variants, states, spacing rules, and behavioural expectations. Teams could reskin or adapt a brand through configuration instead of rebuilding the interface from scratch.

Proteus — final solution visual
Impact

Outcomes

Proteus created measurable improvements across consistency, delivery speed, adoption, and quality. UI inconsistency dropped from around 60–65% to under 20%. Full brand reskins went from 3–4 weeks to under 4 hours. Component adoption grew from under 15% to over 80%, and UI regressions per release dropped by around 67%.

60–65% → <20%
UI inconsistency reduced
3–4 weeks → <4 hours
Full brand reskin time
<15% → 80%+
Component adoption
−67%
UI regressions per release
Proteus — metrics visual

The goal was not to police design. It was to make the best path the easiest path.

Jonathan Pace Leadership · Scale · Systems · Governance
Proteus — final showcase
Takeaway

Design systems are not really about components. They are about reducing decision debt. When the system is structured well, teams move faster, brands stay flexible, and quality becomes repeatable instead of depending on memory, luck, or whoever last touched the UI.

Next project

Pandora's Box

An internal platform concept that gives teams running a large portfolio of brand sites one place to see site health, analytics, documentation, and feedback — instead of a dozen disconnected tools.

Open case study