SaaS · consumer · designing without precedent
A browser with no tabs to copy.
Stack is a spatial, multiplayer browser — the internet arranged in space rather than in a row of tabs. There was no established pattern to borrow from, which is the whole reason the case is interesting.
- ROLE
- UX/UI Designer
- WHEN
- Apr 2022 – Dec 2023 · Amsterdam
- SCOPE
- UI · design system · 200+ screens · real-time collaboration
- TOOLS
- Figma · FigJam · Discord · Adobe CC
- PRODUCT
- stackbrowser.com
/01 — the situation
Tab-heavy browsing produces digital overwhelm.
Traditional browsers offer a fragmented, tab-heavy experience. Users needed a fundamentally different approach to browsing — one that supports focus and collaboration rather than accumulation.
Designing without precedent is a specific discipline. When there is no convention, every affordance has to be taught inside the product, and every novel interaction is a bet you can only settle by putting it in front of people.
/02 — the approach
Space instead of a row.
A spatial browsing experience that reimagines how people interact with the web: multiplayer collaboration, mindful browsing patterns, and a spatial UI designed to reduce cognitive load rather than add to it.
Fly Mode, Realms and multi-profile workflows each needed their own mental model and their own way of being introduced — which is what pushed the work past 200 screens and into a system rather than a set of layouts.
The design carried a fundraising job as well as a product one: an unfamiliar idea has to be legible to investors before it can be legible to users.
Product footage from stackbrowser.com
Teach the model inside the product
With no precedent, onboarding is not a phase — it is a property of every screen.
- CHOSE
- A system that introduces its own concepts as you use them
- OVER
- A conventional tab UI with spatial features bolted on
- COST
- Higher design cost per surface, and no borrowed patterns to fall back on
/03 — the surrounding system
A new model needs more than one good idea.
Cards, Fly Mode and Realms only work if everything around them agrees. The feature surfaces had to teach the model while doing their own job — an assistant that speaks the spatial vocabulary, a shortcut layer that rewards learning, and settings that behave like the rest of the world expects.
This is where designing without precedent gets expensive: every one of these surfaces needed its own onboarding moment, because none of them could assume the user already knew what a card or a realm was.
/04 — outcome
An idea made legible.
- $4M investment secured
- 25,000+ users
- 200+ screens and interactions designed
- Real-time collaboration experience shipped
[FILL] What would be done differently, and the honest failure or reversal from this project.