
How We Safely Landed Our Design System: Lessons from Wolt’s First 9 Months in Orbit
About
Building a design system rarely starts from scratch. More often, you join mid-flight—surrounded by existing patterns, tools, and habits. In this talk, I’ll share how we spent our first nine months evolving Wolt’s design system: introducing a full token architecture, creating a scalable component library, and aligning teams—without disrupting ongoing work. You’ll learn why thinking in orbits, not timelines, helps you navigate inherited systems, and how to design for both today’s needs and the next decade of growth.
Watch the full talk
Watch this WaysConf session, then continue with related talks or explore the current programme.
Talk in brief
This talk recounts the first nine months of building Wolt’s design system from the middle of an established, multi-platform product rather than from a clean component library. It presents a sequence of maneuvers: understand the current landscape, win credibility through tactical fixes while keeping a business horizon, embrace differences across platforms, avoid treating code parity as the only measure of design quality, and use automation and AI where they reduce maintenance work. The conclusion favors a focused starting point, strong collaboration, and continuous adaptation over an idealized system designed in isolation.
Key takeaways
- 01
Start from the system that already exists
An established product carries components, conventions, code, teams, and constraints; useful design-system work begins by mapping that reality.
Watch from 4:20 - 02
Pair tactical wins with a strategic horizon
Fixing icons, tokens, and immediate friction proves value, but the team also needs a view of the business and product direction beyond those tasks.
Watch from 9:21 - 03
Cross-platform does not mean identical
A shared system must coordinate product intent across platforms while respecting native behavior and the different needs of designers and engineers.
Watch from 14:11 - 04
Do not make parity the only goal
Perfect equivalence between design files and code can reduce usability for designers and distract from whether the system improves product work.
Watch from 15:01 - 05
Automate maintenance without abandoning ownership
AI and tooling can help produce documentation and bridge gaps, but teams still need clear decisions, review, and responsibility for the system.
Watch from 20:09
Video chapters
- 0:21Nine months in a new design-system orbit
The speaker introduces a recent practice shaped by iteration, uncertainty, and collaboration.
- 4:20Beginning in the middle of an existing product
Unlike clean-slate tutorials, the team inherits platforms, components, and organizational history.
- 8:59Tactical work with a business horizon
Short-term fixes build trust while strategic awareness keeps the system connected to company needs.
- 13:40Designing across platform differences
Shared foundations are balanced with native patterns and distinct production constraints.
- 15:01Questioning absolute code and design parity
The talk challenges workflows that optimize synchronization at the expense of practical design use.
- 20:02AI-assisted documentation and maintenance
Automation helps a small team address work that lacks dedicated writing and operations capacity.
- 25:40Five maneuvers for a durable system
The main lessons are consolidated into a practical sequence for teams in similar conditions.
- 30:40Questions on teams, adoption, and scope
The closing discussion examines collaboration structures and how a focused seed can begin the journey.
Read edited transcript highlights
These concise notes were edited from automatic captions and checked against the talk structure. They are not a verbatim transcript.
A mature product is not a blank canvas
Most design-system examples begin with a new palette or component library. This team began much later, inside a functioning product with accumulated decisions, multiple platforms, and active delivery. The first responsibility was therefore investigative: understand what already worked, where teams experienced friction, and which constraints could not be wished away by naming a new system.
Watch from 4:20Short wins should point beyond themselves
Icons, tokens, and other immediate repairs can establish credibility because teams feel the benefit quickly. Yet a backlog of small fixes is not a strategy. The design-system group also needs to understand where the product and business are going, so tactical decisions form foundations for the next stage rather than becoming a collection of disconnected improvements.
Watch from 8:59Shared intent permits platform variation
Cross-platform consistency does not require every component to behave or look identical. Each platform brings native expectations, implementation details, and different working needs. The system can preserve shared principles and product meaning while allowing appropriate variation, provided those differences are deliberate and understood rather than accidental drift.
Watch from 13:40Synchronization is a means, not the product
Complete parity between code and design artifacts can sound like the final measure of maturity. The speaker questions that assumption. A technically synchronized library may still be awkward for designers or fail to improve delivery. The stronger test is whether the system supports confident decisions and collaboration across the people who build the product.
Watch from 15:01Begin with one meaningful foundation
The recap encourages teams to understand their position, address tactical pain, work across platforms, use automation thoughtfully, and retain sight of the larger product. A design system does not need to launch as a complete universe. One well-chosen foundation—typography, a core control, or another shared element—can create the first credible orbit for continued work.
Watch from 25:40


