A new CMS for the City of Austin
The problem
The City’s traditional Drupal-based CMS was outdated and mismanaged.
The architecture of the site was department-based, so users had to know which departments were in charge of the service they were trying to access in order to find it, and processes involving multiple departments were very challenging to complete.
There were hundreds of irrelevant content types and outdated pages, making finding what you were looking for very difficult.
Improperly designed content types led to users constantly hacking things and, for example, creating “department pages” for services.
Authors had very few guardrails on the content they created which meant that, given dozens of authors with a variety of expertise levels at writing for the web, hard-to-read content would often get published
The interphase was technical and unintuitive, so departments often assigned IT specialists rather than subject matter or communications specialists to maintain their web content.
The project
Use a content-first, resident-focused approach to create a headless CMS for the City of Austin that would uniquely respond to the user’s needs and support the creation of accessible, multilingual web content.
A much smaller number of relevant content types (above) were created that spoke directly to department and resident needs, and the architecture of the website was changed from department-centric to a services-centered. Guardrails were built in to ensure the use of headings and standardized font colors so that content would be accessible for residents with visual impairments and easy to read for all.
My role
Collaborate in the creation of a services-based architecture.
Help define user roles and permissions.
Set requirements for content types and generate ideas for new features.
Train users on the new CMS and help develop competencies in plain language, web accessibility and SEO.
Work closely with departmental staff to create workflows that would support the creation and maintenance of accessible, up-to-date content across multiple languages.
Transition large amounts of content and stress-test the system.
