NextDoor — UX redesign and design system for a local news appNextDoor — UX redesign and design system for a local news app

Problem

Local news apps live or die on trust, yet the original NextDoor experience buried it. Content was organised by feature rather than by what a reader cares about, and getting to a relevant local story took four levels of navigation. Depth like that doesn't just slow people down — it signals that the app doesn't know what matters, which erodes the exact trust a news product depends on.

The brief: rebuild the experience around trust and community engagement, and make the local story the thing you reach immediately, not eventually.

Research

I audited the existing information architecture and traced the real paths a reader takes to a story they care about. The friction was structural: the navigation modelled the org chart, not the reader. I focused the research on two questions — what makes a reader trust a local story, and how few steps can separate them from it.

• Audited the existing four-level navigation and content taxonomy
• Identified source trust and locality as the signals readers act on
• Found navigation depth, not content quality, was the primary friction

Constraints

A redesign, not a greenfield build, meant respecting the existing content model and the realities of a news product: high posting volume, mixed source quality, and a community layer that had to coexist with editorial content. The redesign had to improve structure without requiring the whole backend to be rebuilt.

Information architecture

The central move was collapsing navigation from four levels to two. I restructured the taxonomy around what readers act on — locality and trust — instead of internal feature groupings, so the most relevant local content surfaces near the root rather than three taps deep.

• Cut navigation depth from four levels to two
• Re-grouped content by locality and trust rather than by feature
• Brought community engagement up as a peer to editorial content, not a buried tab

UX decisions

Trust signals were made structural rather than decorative: source indicators and locality cues live on the story card itself, where the decision to read or skip actually happens. Reducing depth meant the home surface had to carry more, so hierarchy and scannability did the work that extra navigation layers used to.

Design decisions

I built a design system rather than a set of screens — a component library covering story cards, trust indicators, navigation, and the community surfaces, all on a consistent token set. Systematising trust signals as components meant they stayed consistent everywhere they appeared, which is what makes a trust cue believable.

The system was explicitly built for developer handoff: documented components and specs an engineering team could implement and extend without re-deriving the rules.

Technical architecture & handoff

The component system was structured the way it would be built — composable cards, a navigation primitive, reusable trust-indicator components — so the design maps cleanly onto a React component library. This is the design-systems discipline: design artifacts that are already shaped like the code they'll become, which is exactly how a Figma system reaches production without drift.

Results

Navigation depth dropped from four levels to two, putting relevant local stories within reach instead of buried. Trust became a structural part of the interface rather than a claim, and the handoff-ready design system gave the product a consistent, extensible foundation to build on.

Lessons learned

Information architecture is a trust decision, not just a convenience one — in a news product, where content sits tells the reader what you think matters. And systematising signals like source trust into reusable components is what keeps them credible; a trust cue that looks different on every screen stops being a signal at all.