Revitalizing the Symptom Checker
Modernizing and Improving Results
The long-standing Symptom Checker, a hallmark product of the company, had built a reputation for giving broad results. How might we encourage people to input more symptom data so that they would receive more accurate results?
Provide people with dynamic feedback as they follow a simple step-by-step process, data entry is encouraged as needed. Also, update the database to increase the number of more relevant condition matches.
A product in need of a refresh.
The Symptom Checker had not been refreshed in over 7 years.
How might we ...
... modernize to support users on all devices?
... encourage users to enter more data?
...add value and provide ease of use?
Increase ad view-ability
Promote the Physician Finder
Accurate results and related information.
What to do next.
A new technology vendor
Update the body image
Gather more data
Break over multiple pages
A progress meter
Responsive code base
Accuracy ranking of results
Areas of friction
What are the problems we are trying to solve?
Usability Testing and Analysis
"…would be easier if I could just type in [my symptoms]"
Desktop → See report
Most users abandoned the body view in favor of search
Most users had trouble with clarifying questions
Mobile → See report
Half of all user switched to the symptoms list view after struggling with the body map
Many users struggled to enter a second symptom (mobile)
What are the problems and opportunities?
Define needs states / state of mind
Motivate users to enter more data through a results meter
Improve the data entry on smaller screen sizes
Key features to focus on.
Provide users with related symptom suggestions
Present a more prominent and visible Search input, with auto-complete
Break out treatment options as its own section
Display a results accuracy meter to encourage more data input
Show data entry in steps with a progress indicator
Mobile touch-friendly body map that leads to category lists
Let's make solutions!
The Clickable Body
Modernizing the body presented a challenge, the original characters (A) felt cold and, too medical, despite attempts to warm it up. A more realistic approach (B) brought concerns that it would not be identifiable as it was too idealized. Simplifying a "typical" body to outlines (C) but retaining some styling treated it more as an instructional diagram than a physical object.
The development effort to on rebuild and re-map the body map had been underestimated and would jeopardize the launch. A decision was made to move forward with a version that did not rely on the body map.
Launch and Analyze
In the wild!
Results + Learnings
How is it performing?
✓ Page views increase +1
✓ Search field usage
✓ Common symptoms
✓ Medical history entry
A deeper look at the mobile
A decrease in mobile engagement from the previous design led to more user testing.
Users miss the body map!
Data showed a propensity for Users to switch to lists. But, customer feedback received said that they liked the interaction with the body, it was a good way to start, to indicate "where it hurts".
Some terms are too medical!
Our Medical Advisory Board went about the task of editing the terms to be more consumer friendly.