National Aquarium Website Redesign & CMS Implementation

Penn Medicine

User Testing for Leading Cancer Center

TBG & Penn Medicine

When Penn Medicine engaged TBG (The Berndt Group) to redesign the Abramson Cancer Center website and implement Sitecore CMS, we knew we were in for a good challenge: the Penn Web marketing team is a sophisticated digital marketing shop. From the very beginning of the project, the team spoke about testing solutions with real users and not wanting to redesign within an echo chamber. Penn's Web team demanded data-driven design decisions; they wanted proof our proposed user experience would work.

We were thrilled. Many marketing shops have long experience that sometimes translates to assuming the mantle of the user to predict how he or she will behave on a site. But Penn knew the only way to see how users would accomplish tasks on the new site was to test it with real people, and that user tested designs would help the Web marketing team to socialize the solutions across the Penn enterprise.

Penn Medicine is a leading healthcare system in the Philadelphia area, providing world-renown treatment, research, and a broad range of unique specializations. One of Penn’s flagships is the Abramson Cancer Center (also known as Penn Cancer), an institution that is redefining the science and treatment around cancer.

TBG screened and recruited test participants, created a clickable prototype for testing, developed the test script and conducted the testing. Upon completion of the user testing, TBG analyzed the results and delivered a recommendation for some new patient-centered content. We wouldn't have been able to do that without talking to actual patients and their caretakers. We were humbled by the experience of finding out what cancer patients need in a distressing time, and we were honored to be able to help create ways for them to get what they need during that time.

And the Penn Web marketing team was able to proceed in the redesign with the knowledge that they were making decisions based on data. This round of user testing was a win-win for both the client and for TBG.

"Penn Medicine is lucky to have TBG as a strategic partner. Taking into account the complexities of our infrastructure and the sheer size of our organization and existing content, TBG worked with my team to develop a thoughtful plan that will evolve Penn's online presence, and do it quickly. We are positioning ourselves well to be more nimble, efficient, innovative and engaging."

Tanya Andreadis

Director, Interactive Marketing, Penn Medicine

What We Did

Technology Partners

Screening & Recruiting Test Participants

TBG wrote a screening tool with which Penn marketing staff recruited test participants. Penn staff was able to use their long-cultivated relationships with physicians and other healthcare providers to find patients and other targeted participants.

Our group of test participants comprised of:

several current and former cancer patients;
several family members/caretakers of current patients;
one physician who treats patients and is also a cancer researcher;
one nurse "navigator" who acts as a guide to newly diagnosed patients; and
one undiagnosed prospective patient

Yes, our test participant pool was heavily weighted toward people who already have some familiarity with the Abramson Cancer Center. But the entire group, minus one, was over the age of 50 and mostly not routine web users. So although the group wasn't balanced across subject matter familiarity, it did better represent the majority of Abramson's actual users—patients and prospective patients who are middle-aged or older.

Abramson also has a large contingent of patients and prospective patients who are familiar with more complex medical terminology than a typical non-cancer patient or prospective patient would be. People who have been diagnosed with cancer at Penn are more likely to look for and understand complicated scientific content—particularly when the content pertains to their potentially life-threatening condition.

Creating a Clickable Test Prototype

TBG created a clickable grayscale prototype of the proposed Abramson Cancer Center homepage, its flyout navigation menus, and an article page deep in the site. The latter would help us test the local navigation scheme, which had been proved out by creating a prototype of a page at the deepest level of the site.

We used Axure RP to create the clickable prototype. The global navigation flyouts were clickable, as if on a real site, as were the local navigation and a few other elements, such as off-canvas mobile menus.

Because the site we are redesigning for the Abramson Cancer Center is responsive, our prototype had to work at large (desktop), medium (tablet), and small (handheld phone) sizes. We had a prototype of each page at each device size for the users to test. Each test participant completed half the test on a laptop, and the other half on either a tablet or a phone.

TBG used Morae, usability testing software, to record the test participants' voices, faces, mouse trails, and keystrokes.

Developing the Test Script

TBG wrote the test script, which asked the test participants to complete tasks using labels or elements we were least sure about, such as language changes. Penn wanted to convey that not only do they offer world-class cancer care because of the latest technological advances, but that they also offer humane, warm, patient-centered care.

Much of the labeling was changed to reflect this, as well as the overall information architecture. The old site structure organizes content according to departments and other content wells that create silos of disconnected content. The proposed site structure would allow users to navigate by conditions and treatments.

So the test script asked test participants to imagine they have breast cancer. Not a difficult scenario to imagine for most of our test participants, many of whom had been diagnosed with cancer already. The script asked them to navigate to treatment types or symptom checkers, or to get a second opinion. (Getting a second opinion is a common way for prospective patients to come to Penn.) The script also asked that test participants be able to easily and immediately find a doctor and make an appointment, since those are the two primary conversion points on the homepage.

The script contained no leading questions, but we did include open-ended questions at the end of each testing session, to suss out what real users are actually thinking. Through these open-ended questions, we mostly learned about what the test participants thought was missing from the site.

Strong Opinions from Test Participants

Because most of our test participants were actual cancer patients or their family members/caretakers, they had strong opinions about what content they should be seeing, and how it should be labeled. As an aside, it's a lesson in humility to learn how people living with life-threatening illnesses—and the healthcare professionals who treat them—think about the content they need to manage their healthcare and to do their jobs.

In the proposed strategy for the new Abramson Cancer Center site, we stipulated that a user should always know what he or she should do next. One of the ways of realizing this goal was to create new content around getting started as a patient. This content was well received by the patient participants, but they wanted even more of it. They, and the nurse navigator, explained that when you're first diagnosed, or you think you are about to be diagnosed, you want to read reams of content about your condition. And not just consumer-friendly content. They want peer-reviewed scientific content as well. Several patient test participants said the "getting started" section would be the first place they would look upon diagnosis.

An Unanticipated Finding

Another interesting finding from our testing was one we hadn't anticipated: Users want to know whether their insurance is accepted right off the bat. Of course they do! When they look to Penn Medicine for world-class cancer care, they also want to know if there's a deal breaker—whether their treatment will be paid for by their insurance. And they want that deal breaker to be known before they go too far in their journey with Penn.

The rest of the testing results were unremarkable: A few problems finding things, a couple of people not able to tell what section of the site they were in, and so on. For the most part, the testing confirmed the design decisions TBG and the Penn web marketing team had made.

SafeNet