Project 4: Information Architecture & A/B Testing

ROLE: UX researcher 

Led planning, programmed studies, analyzed, reported, presented to executives

METHODS: Generative research, heuristic evaluation, card sorting, unmoderated A/B prototype testing

OUTCOMES: Identified usability issues, created a taxonomy, delivered insights & recommendations

TOOLS: Microsoft Office, UserZoom 

SKILLS: Strategy, research, data analysis, reporting, stakeholder education, executive presentation

While conducting in-home contextual read-along sessions would have been ideal, the client's budget did not allow for that. Therefore, we conducted remote read-along sessions so the participant could still show us what they were doing while in their home environment. 

During the engagement, I used the following methods:

• Overall research strategy: I worked with another researcher to create the overall user-centric research strategy for the discovery, information architecture, design, development, and launch phase of the project to ensure the project team kept the focus on the user experience. 
• Heuristic evaluations: While we recommended conducting a baseline usability study during the discovery phase to understand the current successes and challenges of the site, the project team ended up vetoing it due to the new budget and time constraints they placed on the project. As a result, my partner and I conducted separate heuristic evaluations on four key areas of the site using a standardized scorecard so we could each evaluate the sites in a standardized and measured way. 
• Round 1 card sort: I started with an open card sort on two of the audiences to understand how users were thinking about the organization of the site and to determine the direction for the taxonomy. The project team decided to keep some of the current jargon and were overly specific in their screening requirements leading to a challenging project. As a result, we struggled to get enough users that met the narrow requirements and those who did complete it had difficulty with the labeling used on the cards. The results from the study ended up being inconclusive.
• Round 2 card sort: After presenting the results to the project team, they opted for another open card sort with revised labels and more realistic screener criteria. The round proved more effective and allowed me to provide a recommended taxonomy.
• Unmoderated A/B prototype testing: The team put together two low-fidelity prototypes for three audiences using the recommended taxonomy and wanted to understand how users would navigate to the content, whether or not they would use search, and which of the prototype layouts was preferred. I decided to conduct an unmoderated A/B study in order to get larger numbers of participants to best understand their navigational behaviors and preferences.