Signal › Contributors
Léa Dupont
I studied interaction design at ENSCI in Paris and then made the classic French designer mistake: I moved to San Francisco. I spent two years at a design agency in SoMa doing contract work for startups that were mostly spending their seed funding on branding they didn't need yet. I designed seven logos, three brand identities, and one pitch deck that helped raise a Series A. (The company pivoted six months later. The logo survived.)
I moved back to Paris in 2018 and joined BlaBlaCar, the ride-sharing platform, as a product designer. BlaBlaCar was an interesting design challenge because the product had to work for two very different user groups: tech-savvy urban millennials and older suburban users who were primarily motivated by saving money on fuel. The same interface had to feel modern and trustworthy simultaneously, which is harder than it sounds.
After a year of shipping features, I realized that our design inconsistency was slowing us down more than any individual feature decision. Three product teams were building the same button three different ways. I pitched the leadership team on building a design system. They said yes. I spent the next two years building Plasma, BlaBlaCar's design system, from the ground up: token architecture, component library, documentation, adoption metrics, the works.
In 2021, Datadog recruited me to lead their design systems team. Datadog was a different universe: 1,000+ engineers, dozens of product teams, and a product surface area that included dashboards, monitors, logs, APM, security, and about 15 other products that all needed to feel like one coherent thing. Building a design system at that scale isn't a design problem; it's an organizational problem. You're not designing components. You're designing governance models, contribution workflows, and incentive structures that make it easier for teams to use the system than to ignore it.
The thing I learned at Datadog that I now write about constantly: a design system is a product. It has users (designers and engineers). It has adoption metrics. It has a roadmap. And like any product, it fails when it's built based on what the team thinks users need rather than what they actually need. The number one reason design systems fail isn't technical. It's political. Someone builds a beautiful component library, ships it without talking to the teams who'll use it, and then wonders why adoption is at 30% after a year.
I joined Signal because design systems discourse is overwhelmingly tactical (which Figma token plugin should you use?) and almost never strategic (how does your design system affect your organization's velocity?). I want to write about the strategic layer.
I live in Le Marais, Paris. I paint watercolors on weekends, I have strong opinions about typography, and I believe that Helvetica was a mistake. (I'm mostly kidding. Mostly.)
Experience
- Design Systems Lead, Datadog
- Design System Creator, BlaBlaCar (Plasma)
- ENSCI – Les Ateliers, Paris