Proving the design system pays
How a shared design system became one of the clearest routes to revenue, not a cost of doing business.
The question every design system gets asked
Design systems are usually budgeted as a cost of doing business: useful, worthy, hard to measure. When budgets tighten, that framing puts them first in line. I wanted ours to be judged on what it does for customers and the business, so the work had to be measurable from the start.
Speed is the lever
A shared system means most of our customer experiences are built from the same components, and design to build is around three times faster as a result. Faster builds mean more experiments. More experiments mean we learn sooner what customers actually respond to.
- of customer experiences built from the same components
- 90%
- faster from design to build
- 3x
- typical landing page build
- 2 weeks → 4 days
Where it shows up in revenue
The clearest proof came from in-app renewals. We migrated the renewal journey onto the design system and shipped it as an experiment. It converted 38% better than the version it replaced, an uplift attributed to the migration itself, and it now serves every app user.
The same foundations carried the rebuilt web self-serve platform, which cut customer contact by 30% and web errors by 88%.
- uplift in tariff renewal conversion in-app
- 38%
- reduction in customer contact
- 30%
- fewer web errors than the legacy experience
- 88%
Measuring the system itself
Claims about adoption are easy to make and hard to trust, so we built the evidence. Element Analytics tracks design system adoption across our codebases. Alongside it, an audit loop scored all 127 components across six dimensions and turned the findings into engineering-ready tickets on a product team's board, with fixes verified within the week.
That changed the conversation with leadership: system health became a dashboard, not a debate.
What I'd tell another team
Pick the commercial journey where consistency and speed matter most, migrate it properly, and measure it as an experiment. One clean result moves the conversation from "nice to have" to "where next".