Senior UX Designer
The healthcare dashboard serves as the home screen of the platform and the initial communication center for users. Through feedback and analytics, we saw that the dashboard's engagement was extremely low, and the communication channels were not effective.
The dashboard's UX needed to be optimized to allow for customization for different types of users. The communications needed to be streamlined, and support channels needed to be reimagined.
Empathize
Define
Ideate
Prototype
Test
One of the first things I did was to reach out to users and interview them to determine who would benefit the most from a healthcare admin dashboard. Luckily, there was an established list of friendly organizations I could set up 30 minute zoom meetings with to discuss their experiences and pain-points.
I discovered that large organizations with over 50 users, servicing over 100 providers with in-house IT support and sophisticated software would benefit the most. An example of this type of organization would be a large chain of medical clinics.
The next step I took was to facilitate a series of discovery workshops with key stakeholders to see what problem we would like to solve and what they believed success looked like. We came up with a set of goals and objectives and a shared problem statement.
Improve Communications
Dashboard components and graphs should be configurable
Establish Task Lists to Increase Efficiency
Create Effective Help and Support Channels
How might we make the healthcare dashboard more relevant and useful to meet the needs of all our different users, and add value by leveraging new technology?
Next, I set out to work with a cross-functional team to come up with as many ideas as we could to solve the problem. From there, I used an affinity diagram exercise to narrow down the ideas to two or three. In Miro, I started an inspiration catalog with an inventory of products that had some of the features we had discussed as a team.
Before jumping into a prototyping tool to create high fidelity prototypes or mockups, I like to begin with simple wireframes. I used feedback from users to help me quickly sketch out ideas. I found that users did not like a lot of change, because they are measured by their productivity. When things are moved around or changed, they have to take time away from tasks to learn a new system. Knowing this, I tried to work within the established information architecture set by the application's homepage.
I then showed this rough sketch to users to determine if I was on the right track. Using feedback from these informal check-ins, I moved on to a high fidelity mockup using Figma. Below is a quick tour of the dashboard.
I ran two sets of tests against the prototype. One was moderated. I was able to meet with about 30 organizations and facilitate a structured usability test. The second was unmoderated. I used gainsight to recruit users of the current application and send them links to the prototype. I then used hotjar to set up a series of tasks for users to interact with and create pre and post-surveys.
Once I felt confident about the design based on user feedback, I worked with the development team to create a beta version and piloted it with a very small group of users. I worked with key stakeholders to determine success criteria for the pilot.
The healthcare admin dashboard met all of the success criteria set by keystakeholders. In fact, it received only 6% negative feedback from beta users.
In the end, the client decided to postpone the launch of the new dashboard and go a different direction. However, I learned a lot of great things from this project.
Users are resistant to change. The more information we can provide ahead of time to users the better. Also, little contextual guides, tips and tricks in the UI lessens the learning curve.
The defaults matter. I've heard that 90% of users stick with the default settings and even with a highly configurable dashboard we saw the same pattern of behavior. Nailing down what is most important to users to initially show is key.
Personalization costs! It can cost in performance, in development hours, and in data mining and storage. We had to make some trade-offs to accommodate the rising cost of personalizing different aspects of the dashboard.
Say hello.
teresawyattlopes@gmail.com
Dribble
Behance
Instagram
LinkedIn