Preparing Tasks & Script

We drafted the initial testing overview, including task goals, pages needed, and alternative flows. Once we had a draft of the initial test, we sent the script and a preview link to the client for revisions to tasks and for feedback on language. Our biggest challenge while preparing the tasks and script was ensuring that the agreed upon tasks and language was general enough to allow for users who might use different terminology than the business used. To ensure our task language didn't skew the test results, we focused the tasks around common user goals instead of site sections, intentionally using language different than the design.

Building & Populating the Test Pages

Next, we built out all the necessary wireframes and pages to complete the path, including any alternative paths identified in the task preparation. This required two main steps:

• Content Mapping - Visualizing current site content with new site components. Client provided word docs with all content for priority pages, and we did the dirty work.
• Content Population - Creating wireframes of the pages necessary for population and linking all necessary elements for the test together into a working prototype using Axure.

Recruitment & Data Collection

We used UserTesting.com to recruit participants for the test, conduct the tests in an unmoderated fashion, and collect data for our analyses. Our test included screening criteria to get as close to the potential user base as possible, such as location (Maryland only) and occupation (Small Business owners — who likely purchase Worker's Compensation insurance) Before setting the test live on the UserTesting platform, we ran a pilot version of the test with members of the internal team to understand how people would interact with the prototype and understand any limitations of the platform, and we briefed our real participants accordingly based on these learnings. We collected data from 6 users to identify what potential issues might exist in key user flows

Turning Research Findings into Insights

To analyze the test data, we completed the following steps:

• Watched all the videos, annotating with notes in an Excel document with timestamps for ease of access.
• Scored each task using the above Excel document — users who completed the task first try received a score of 1, with each subsequent attempt receiving half the score of the previous trial.
• Summed up the scores for each task across the entire group to determine which tasks were hardest for the test group
• Scored the self-report data on ease of use and other subjective measures of usability.
• Documented all recommendations to guide functional requirements for the changes coming out of testing.

Based on task success, our biggest recommendation during the first usability test was to increase visibility of the filters/facets in the Safety Center to help users find articles and trainings that are relevant to their needs more easily.