Inspire Search Redesign
Inspire is the world’s largest online health community, patient engagement and real-world data platform. More than 10 million people each year come to Inspire to learn about their conditions and find peer support. Inspire also works with leading biopharmaceutical companies and academic researchers to provide real-world evidence and insights about the patient experience, connect patients with life-advancing clinical trials and enable medical breakthroughs.
TL;DR Before and After
We did an overhaul of the site's general search feature to focus on usability, relevancy, and customization.
Problem to solve
Why search?
The previous search feature was not intuitive and caused a lot of confusion for our membership. Our community moderators were fielding questions and complaints about the functionality and usability of search. We set out to revamp the Search feature to create a scan-able, customizable and informative experience.
-
Technology issues: The search engine would consistently return information that was not relevant to the user’s search term.
-
The filter and sorting options caused confusion for members and anonymous visitors.
-
Items displayed in carousels were often overlooked.
Collaboration
My role
I served as the UX design lead on this project.
I worked closely with a UX researcher to conduct user testing and establish a baseline knowledge from previous relevant research.
After the wireframes were completed, I collaborated with a UI designer to bring the mock ups to life and fit within our branding guidelines.
Before hand-off to development, I liaised with product managers and stakeholders to make sure the product design was feasible and helped define requirements.
Process
First things first
I conducted a competitive analysis comparing major search engines against those of similar forums. I narrowed down specific features that were important to include, making a priority matrix and measuring against requirements provided by product management.
We gathered any research that was previously done that might be relevant to this project as well as interviewed stakeholders to make sure all of the requirements were defined. Because there had to be a shift in the back end search engine, we worked closely with engineering to identify any constraints or barriers.
Ideation
Wireframing
I created several rounds of wireframes, flushing out new ideas while balancing both product requirements and business needs. Product management was involved through iterations to provide feedback and make sure we were within engineering's scope.
Something we were interested in exploring was surfacing replies within the search results. Because we are a health forum, a lot of valuable information is found within the comments sections of posts. During testing, participants responded positively to this idea. Ultimately this feature did not make it into the MVP, but there are plans in product roadmap to add it as an enhancement.
Research
Concept user testing
Once we optimized the wireframes, we put them in front of users in an unmoderated user test. Using the testing software Userlytics, we recruited a panel of 16 participants: eight Inspire members and eight users who are not familiar with our product.
This test provided us with invaluable insights including:
-
Sorting and filtering components were a defining feature for a successful search. Test participants wanted more options to narrow down their search.
-
The ordering of sections was important for users to feel like they are getting relevant information instantly. We found that the ideal order was posts/replies, communities and then articles.
-
The “search within communities” dropdown caused a lot of confusion due to wording and placement.
We iterated on the wireframes taking the recommendations provided through the research such as, but not limited to,
-
Post restructuring so that important details were prioritized
-
Changed community list to be vertical for easy scanning
-
Reconsidered labeling for sort and filter menus
Takeaways
-
While we tested the low-fidelity wireframe to prevent duplicative UI work, it became clear that the clickable prototype was confusing for test participants. In the future I am inclined to run tests with mid-fidelity to high-fidelity prototypes for ease of use.
-
It is crucial to establish comprehensive requirements with the product team before beginning design work to maintain a realistic scope.
