Information architecture
Highlights: information architecture, data analysis, research, taxonomies, navigation, content operations, content migration, content governance, user surveys, card sorts
TL;dr: As part of a client’s site redesign, I restructured one of the most useful sections for client employees to match real user behaviors, and supported the client in rewriting and migrating content to improve the user experience.
People problem
Background: My employer was contracted to redesign the website and create a member dashboard and enterprise content management system for a province-wide union in Alberta, Canada. I was the lead content strategist on this year-long project.
In the discovery phase of this project, I noticed through analytics that the discounts offered to employees were one of the most popular areas of the site.
There were hundreds of discounts available, but the location filters were inconsistent and had a ton of overlap, and the category filters weren’t mutually exclusive.
This navigation made it tough for employees to complete an essential task.
Sample analytics from the discovery report. Client name redacted.
My strategy
To help employees find the discounts they were looking for, I wanted to understand how they used the site to find them. With that information in hand, I’d use information architecture practices and user research to allow users to create groupings based on their own mental models. Finally, I’d work with my design and development team to redesign the discounts based on real employee wayfinding.
What I did
To learn more about how users look for discounts, I surveyed site users on what search terms they use when looking for deals.
I allowed users to manually input their preferred search terms
Then, I put those survey answers through a text analysis tool, looking for patterns
Sample survey results from my client report. Client name redacted.
I audited the categories on the current site, looking for patterns and places to merge or create mutual exclusivity
Card sort activity participant screenshot.
Card sort activity similarity matrix to show common groupings.
I created a card sort exercise to allow users to create their own groupings of discounts, based on their own mental models
In analysis, I looked for patterns in the groupings of content to crowd-source associations and category titles
Based on what I learned from user research, I re-wrote the location and category metadata to better support how employees were searching for discounts
I worked with the client to support them in rewriting their discount content, so that the discount title clearly stated what discount they would receive and gather and upload brand images for quick recognition. As part of the full website and enterprise content management system overhaul, my support included:
A workshop on how to write for web with all their content authors, focusing on accessible language at an appropriate reading level, the psychology of how people read online, and SEO
A content workflow planning workshop to learn and rebuild their content governance process, including how content needs are identified, created, edited, published, and maintained
A content operations workflow to aid content migration through GatherContent
Content migration management, from timeline planning, to roles, to ad hoc support
I worked with my design and development team to redesign the discounts landing page to display discount page titles (instead of business names), category tags to easily distinguish between businesses that offered multiple discounts, and to show brand images
Before and after
Discount locations on the client’s old site.
Discount locations on the redesigned site.
Discount categories on the client’s old site.
Discount categories on the redesigned site.
The results
We haven’t yet been able to gather a large set of data on users’ ability to find discounts more easily, however the ad hoc feedback we’ve heard so far has been overwhelmingly positive.