The Problem:
Research, UX, visual design Google Analytics, surveys, interviews, as well as heat-map recordings showed that the Provider Finder/Find a Doctor tool was the most frequently used tool out of all Bright Health products. It was also the tool that garnered the most number of complaints from our users and led to high call volumes at our call centers. I was tasked with streamlining the search and browse functionalities by stripping away all unnecessary interactions to make the experience of finding a doctor faster and easier.
Target Users:
Health insurance buyers as well as health insurance brokers and internal call center representatives
Pain Points:
Inefficient information architecture: The Provider Finder tool was hidden under the Find a Plan dropdown, making it hard to discover. Card sorting exercises revealed that most users did not expect to find the provider finder tool under the plan section at all. The re-design was an opportunity for us to realign with user expectations.
Distracting interactive elements: The secondary nav allowed users to switch the insurance plan they were browsing. However, we found that most users had already made that decision prior to landing on this page and therefore rarely interacted with this navigational element. The way the sticky nav was built to vanish upon scroll often distracted.
Information hierarchy issues: the map occupied 2/3rd of the screen real estate and drew users' attention upon landing, but offered limited functionality and interaction. The only actions a user could perform on the map were zoom and drag, neither of which updated the search result automatically.
Three independently scrolling sections on the results page made it difficult to focus on the task at hand. During user testing sessions, users complained about not being able to guess where they should focus their attention.
Brainstorming
The solution
Iteration 1:
User interviews had revealed the need for simplicity so customers could find what they needed and either make a buying decision or continue browsing our site. Our first hypothesis was to remove the map altogether, and let users look at the results in the simplest form possible.
What we learnt through user testing and interviews was that even though the tool was simpler to use, users were copying the address and pasting it into Google Maps to determine how far the doctor was. They were quick to drop out of the tool after doing the manual copy and paste action 4 to 5 times.
Iteration 2:
We included a search by zip code functionality for users to find doctors close to their location and designed a simple map that showed a pin for each doctor.
In our tests users pointed out that the distances we were showing were inaccurate. We quickly learnt that zip code borders are fairly irregular and while our tool calculates distances from the center of the zip code, that might not be what the user was expecting.
The MVP:
I spent a week working closely with the engineering team to find a solution that was reliable and simple enough for the team to implement. We decided to add a "search by street address" capability along with the zip code input. It allows for more accurate distances to be shown, greatly reducing the frustration that users faced. We also changed the way the search fields were laid out so they were easier to use for regular as well as power users.
Mobile solution:
We separated the results into 2 views: list view and map view for smaller screen sizes. It allowed us to greatly simplify the interaction with the map and gave us enough real estate to show results within the map view itself.
While in the map view, users can edit their search by simply moving the map.
