A concept app from zero

Inside Virgin Media O2, customers' products were split across two brands with no single surface. As a venture-concept project, I took a proposition, one app presenting personalised offers and products from both brands, from zero to a credible, testable concept.

Research first

The process was agreed before the work: definition, visual identity, motion, design system, test and iterate, evolve. The design decisions were grounded in research from the start, built on five persona "tribes", from Mobile Minimalists to Digital Newbies to Savvy Customisers. Each was a full artefact: goals, frustrations, problem-solving journeys, confidence and self-serve scales, and verbatim voice. Personas as inputs to design decisions, not deck decoration.

An identity built from scratch

A variable font, chosen with an explicit static-versus-variable rationale; a nine-tier type scale with weight, size and hierarchy rules per tier; a colour system that was accessible by construction, with every foreground and background pairing contrast-checked before use; and a bespoke icon language of several hundred icons. The design language itself was run as a living, tracked artefact (jobs to be done, documented decisions, managed evolution), not a one-off brand PDF.

Built to graduate

Because my background is design systems, the concept was built like one: tokenised and componentised, so if the proposition graduated to production, the design work would graduate with it rather than be redrawn.

Outcome

A complete research-to-design-language concept product, and proof that systems discipline and expressive product craft are the same designer. The persona practice carried forward into my later archetype-driven design work.