Back

South Bend

Unifying content architecture for a city government

Lead Designer & IA  ·  2025  ·  Civic Design  ·  Content Strategy

Publishing flow diagram

Skills

Information Architecture  ·  Content Strategy  ·  CMS Design  ·  Facilitation

2

City domains unified

7

Content types designed

5

Staff roles mapped

The City of South Bend’s digital content was split across two domains, making it hard for residents to find what they needed. Behind the scenes, the problem was just as acute — staff across multiple departments were creating and managing content without clear ownership, consistent structure, or shared standards.

The city hadn’t yet committed to a CMS. Part of the work was helping evaluate options and make the case for one — ultimately landing on Payload, an open-source platform suited to the city’s needs for structured content and AI readability.

What it needed was a foundation — a unified architecture, a content system, and a team equipped to carry it forward.

Mapped the existing system and its gaps

Working alongside a service designer and in close partnership with the product manager and web manager, I led discovery interviews across city departments — talking to subject matter experts, department editors, and content owners to understand how content was actually created, reviewed, and published. What emerged was a picture of fragmentation: content owned by no one, workflows that varied by department, and a publishing process that put too much on too few people.

Designed the information architecture

We inventoried both page types and the types of information that lived on each page — services, permits, licenses, events, departments, reports. That inventory became the foundation for a unified IA spanning both city domains. I mapped content into linked Notion databases, treating them as a living prototype we could test and refine with stakeholders before anything touched the CMS. This let us validate structure quickly without committing to code.

Content ownership model Permission variations

Prototyped in the real CMS

Having selected Payload, we moved into prototyping directly in the city’s CMS. I designed content templates for each of the seven content types, with structured fields that would make pages consistent, scannable, and readable by AI systems as well as humans. The content editing interface was designed for non-technical staff: clear field labels, helpful guidance copy, and a publishing workflow that gave the right people the right level of control without creating bottlenecks.

CMS interface Form sections and states

Built capacity on the team

In parallel, I ran a series of lunch-and-learns with the team — Figma basics, design libraries, how to work with files collaboratively. They’ve since run benchmark usability tests and applied job mapping independently, building directly on what we worked through together. The goal was always for the methods to outlast the engagement.

The engineering team owned a wide surface area across the city. Bandwidth was real, and implementation wasn’t going to happen on a design timeline.

That shaped how I worked. Every deliverable was built to be picked up without me — the IA documented, the templates structured, the CMS prototype functional. The goal was to hand off something the city’s designers could own, or that engineers could build from directly, without me in the room.

The project produced a unified information architecture across two city domains, seven content type templates, a role-based governance model, and a working CMS prototype in Payload. The system was designed to serve both residents navigating city services and the AI tools increasingly used to surface public information.

The deliverables are complete and ready to build from. The work holds up on its own.