The biggest challenge was to build a sustainable usability testing system from scratch. At the time when I joined the team, there was a full set of un-tested UI, and no actual users on the platform.
“How might we test the usability of the app using UX research methodologies that are most suitable for the current product development stage?”
I started by asking myself the question "What are the goals I'm trying to achieve?" From the discussions with PM, I listed the main business goals first, and then came up with design goals.
I followed that specific order because business goals provide a guideline for design goals, and can make sure I'm not designing for the sake of design itself.
Then I moved on to the research of available UX research methods and identify which ones are most suitable for the current product development stage, and most valuable for meeting the goals. Ultimately, I decided on the following UX research methods:
Upon settling with the UX research methods, I flashed out the entire testing process of initial screening, recruiting, test scheduling, emailing instructions and confirmation, conducting interviews, recording test results, etc.
The flow chart below is a visual representation of the testing system I presented to my PM and other team members to help them understand the process and each case.
Among all the available research methods, I used three that were most suitable for Mayv's current product development stage.
For the prototype testing, I designed two activities to get insights on how users actually interact with the product:
Why did I use these methods?
Task-based Figma prototype testing - To observe how real users interact with the app, therefore find out any design defect that causes confusion or difficulty to use.
Card sorting - To gather insights on whether the current solution Mayv provides is valid and to find new solution ideas for future development.
In order to keep track of the users and log corresponding testing stage progress and results, I created a user relationship management system that records each person's name, contact information, source of connection, interview status, testing result rating, and other relevant details.
With this chart, I could quickly identify who are the users with the best testing results and find relevant information as needed. It also helped me to keep a record of each person Mayv could contact for future tests.
After conducting 32 interviews during the first 2 months with 2 to 3 interviews per week on average, I summarized the following insights from the test results:
From the insights, I soon realized these are accessibility issues with the original UI. For Mayv's target audiences, which are elders, they need an app that is easy to read and browse, and has icons that have clear labelings of what they are.
Therefore, I suggested the following design strategies to improve the usability:
After 19+ rounds of iterations with the external agency, we finalized the UI for the app launch in November 2020.
Drop me a message at siachang.work@gmail.com