City of Bend: Navigation Improvements
The City of Bend is a municipal government that provides a wide array of services and information for the community—navigating its website to find these services was a challenge.
Roles
Project Owner / UX Designer / Researcher / Content Manager
Team Size
2
Platform
Web
Tool Stack
Optimal Workshop / Google Analytics / GovAccess
Background
The City of Bend had been receiving anecdotal feedback—internally and externally—regarding the confusing nature of the site’s IA. It was decided to dig deeper to identify the top tasks site visitors were looking to perform, and the pain points they were experiencing.
Results
Since implementing the new navigation, more people have been able to find the pages they needed and do so without having to click around as much as before. We have also included multiple ways to navigate services: with the addition of an action-oriented “I want to” menu, users can choose to browse for services and information by action verb— “apply”, “register”, “learn”, etc., or start with a drop-down showing the top services as identified through the top task analysis.
Learnings
Due to limitations of our current content management system, I was unable to make some UI adjustments at this time that I believe would further improve the user experience. I want to replace the primary, top navigation with a hamburger menu. City websites hold a multitude of disparate services. When a user is on any given page of our site, they don’t need to see every other service—or type of service—the City provides. For example, if someone is looking for information on how to turn their home into a short term rental, they do not need to see a top navigation that includes “Events”, “News”, “Government”, etc., as these categories don’t pertain to the service and information they’re looking for. Instead, this valuable real estate should be replaced with a page-specific menu.
View the work at bendoregon.gov.
Research & Roadblocks
To update and improve the City's website, I conducted usability testing within the Bend community. This user research included a top task analysis to identify the most common tasks visitors performed on the site, tree testing to evaluate menu labels and categories, and in-person user interviews to obtain feedback for more qualitative data.
Top Task Analysis
To begin, I conducted interviews with internal stakeholders and frontline staff from every externally facing City department. These staff members shared the top services and information their respective departments provided. With the help of Google analytics, I was able to boil down the top services and information provided by City staff to a list of the top 15 tasks for site visitors.
User Interviews
To observe how people navigated the site in real time, and to get candid feedback and impressions, I set up a table with a laptop at various local coffee shops around town. I then offered to buy patrons a cup of coffee in exchange for 10-15 minutes of browsing the site in search of the services and information identified as the top tasks. Common themes expressed included:
Too many menu options.
Too many clicks to get to important information.
Tree Testing
I conducted tree testing with over 250 participants to determine the baseline success rate for community members finding the right pages on our site when presented with the top tasks. Key findings included:
Ambiguous IA due to sitemap matched to the City’s internal organizational structure—i.e. services and information were located under the individual department or division that performed the service, rather than a task-oriented UX.
Confusing categories that needed to be narrower. Several tasks could have reasonably been found under multiple categories, leading users to guess which one was correct, increasing the time and effort it took to find the correct page—if they did at all.
Analyzing the data we received through the baseline tree testing, along with findings from user interviews and competitive analysis, we designed and tested a new information architecture. This new IA had an overall success rate 14% higher but only a modest 2% increase in directness (i.e. users backtracked only slightly less than before). We then tweaked the IA and conducted another round of tree testing. This second information architecture performed much better, with an overall success rate increase of 21%, and a directness increase of 19% over the baseline IA.
Implementation
Once the IA was determined, I began the process of implementing the new navigation. This involved URL mapping to prevent broken links, and required me to build a new Services landing page and an About Bend page to house content that lived on multiple pages. This page now serves as a high-level introduction to Bend, the location, and the City of Bend, the local government.