Information Architecture Redesign
Sharon Public Library
Project Background
OVERVIEW
Sharon Public Library, a small library in Massachusetts, had a significant amount of online resources in addition to the library catalog. "While there is an abundance of information, it can be daunting for our users to FIND that information."
PROJECT GOAL
Rework the content from the current library website so that patrons can easily accomplish their high-priority tasks
MY ROLE
UX Designer
Information Architecture, User Testing, Wireframing
CLIENT
This project was an assignment as part of my MS in User Experience Design.
Original Site
Two main information architecture (IA) issues hindered users from accomplishing their basic library website tasks.
First, the website used three organizational schemes: user, topic, and task. Content was located in multiple locations, making it difficult for users to find, and to keep updated.
Second, users had to navigate to a completely different website to access the library catalogue.
Research
User Research
Librarians were interviewed about who library patrons are, and what tasks they want to accomplish. Additionally, a literature study of articles and studies gave insight about online library patrons. From this combined data, three personas were created in addition to a task list showing which personas performed each task and the priority of the task.
Content Analysis
All at the same level
Sharon Public Library's existing content was examined. The content analysis showed the page structure, content description, content that needed created, and location on the redesigned site (or if it would be deleted).
The content analysis uncovered a third weakness in the original IA. Although the pages were visually presented in hierarchies (parent-child relationships), the URLS showed that all the pages were structurally on the same level. Additionally, much of the library's content had not been updated in years and needed to be culled.
Design
Classification Scheme
Just one
After discovering what content users needed and what content was already on the website, the new classification scheme was designed. The content was now organized by topic, with one exception. Events and classes were a database so users could use filters to find events. The new organizational scheme was not only easier for patrons to navigate, but easier to keep content up to date.
Sitemap
Structural, not visual
The sitemap shows the hierarchal relationships between the pages. Unlike the original website, this hierarchy is deeper than a visual presentation and reflects the actual structure of the new site. Content that patrons need for high-priority tasks is located on upper-level pages for easier access.
Testing
Tree Study
Six participants who had used a library website in the last year participated in an unmoderated tree study on Treejack. After reading a scenario, participants indicated where they would look for the information needed. The scenarios focused on testing the high-priority tasks identified during user research.
During the test, participants saw the labels from the hierarchy as a series of drop down menus (a tree).
Participants selected the label where they expected to find the information for the task in the scenario.
Analysis
Data from the tree study included participant's first clicks, paths taken, and if they succeed or failed to find the content location. After analysis, most of the labels directed participants to the correct location, but two pairs of labels caused confusion.
In all the scenarios targeting content in 'Location' or 'About,' participants looked in both locations equally. There was not a strong difference between these two categories in users' minds. 'Resources' and 'Services' were also interchangeable to participants.
​
In the next iteration, 'Location' became a child of 'About.' The rest of the confusion was solved by creating a single 'Resources & Services' category so users no longer had to choose between two similar options.
First Click Study
Low-fidelity wireframes were designed for screens where the high-priority tasks were performed. These screens were tested with an unmoderated first-click test on Chalkmark. Four new participants were presented with the same scenarios as the tree study, and then clicked the area of the wireframe where they expected to find the information.
Heatmaps showed where each participant clicked during the test. Most scenarios had multiple correct paths and participants were generally successful finding a correct path with their first click, so no changes were made to the wireframes.
Takeaways
Unmoderated testing
During this project I learned how to design, conduct, and analyze the data from remote unmoderated tests. One thing I would change is to write slightly longer and more narrative-based scenarios. The scenarios from this study did not lead the participants to the answer, but were a little abrupt.
PROJECT RESULTS
The new IA for Sharon Public Library removed outdated content, organized it in one overarching system, and matched the visual presentation with the structure of the website. Test participants were able to find information - frequently with their first click.
TOOLS USED
Miro
Treejack - Optimal Workshop
Chalkmark - Optimal Workshop
wireframe.cc
