Overview
Zocdoc is a medical appointment booking site that disrupted the long-standing tradition of calling doctors’ offices to set up appointments. Users can search for providers of every type of healthcare, but some areas of care are harder to navigate than others, because there are many specializations. Zocdoc’s search experience can put the onus on the user to know what kind of specialist they need. Over 6 months in 2021, I worked on search experiences aimed at matching patients to appropriate providers and improving their confidence to book with them.
What I did:
Wrote copy for a guided search flow for mental health care, which performed very well upon launch
Designed a similar experience for orthopedic care, in which I gathered background from providers, informed and analyzed qualitative interviews with patients, and used the research insights to polish the copy
Created and maintained guidelines on how to strategize and write content for guided search experiences
Mental healthcare pilot
Mental healthcare was a prime candidate for a pilot because of its many specialties and modalities. Someone who enters “therapy” will see a mix of providers including psychologists, psychiatrists, psychotherapists, and MFTs. By having users answer a few questions after entering a search query, we could present more relevant results, saving them time and reducing the risk of them booking with the wrong type of doctor (i.e. a psychotherapist when they are looking for a prescription), which leads to a lot of wasted time for both parties.
I entered the picture as the MVP was being firmed up, and smoothed out the copy for launch. Users enter the flow after entering basic search criteria on Zocdoc’s homepage:
Click through to see the experience before I worked on it (and use “r” to return to the start):
In my revisions, I focused on setting expectations, making it conversational, and keeping the copy as streamlined as possible:
Key decisions
There was debate about including the intro screen. I felt it was important for setting expectations and enticing people to continue. We could always reconsider if there was a high bounce rate. I kept the copy direct about what to do and why.
I proposed only providing feedback in cases where it’s educational. If they select “Prescription”, they see “We’ll show you psychiatrists and other providers who prescribe medicine”. They won’t see “We’ll show you providers that offer video visits” when they select “Video visit”.
To help educate on how insurance works with mental healthcare, I removed the subcopy from the “What is your insurance?” screen and used the feedback pattern to give users information relevant to what they entered.
Results
The mental health flow launched to great success:
72,000+ people saw the intro screen during the pilot and 78% clicked “Continue”, 17% opted to just see all the search results, and 6% bounced
90% of users completed the entire flow and every screen after the intro screen had a 95% and up completion rate
3.6% more users booked after going through this flow than those who didn’t
Cancellation rates were lower for users that booked through this flow
Creating guidelines
Knowing we would be making more experiences like this, I took the initiative to create guidelines for guided search flow content. I covered how to apply content strategy at the outset and provided tactical advice on how to craft questions and apply an appropriate voice and tone. Scroll to view an excerpt:
Next up: orthopedic care pilot
Orthopedic care is complex and very specialized. It is a huge field, encompassing any provider who treats muscles and bones. Many people don’t know the term; they just know that their [insert body part] hurts.
The core Zocdoc search experience isn’t optimized to enable a hand surgeon to only get patients who need hand surgery, and this was actively deterring orthopedists from using the platform. Practices were taking on big monetary and administrative costs by screening Zocdoc patients before visits to make sure they were booking with the right kind of doctor.
Designing the flow
Enterprise orthopedic care practices told us the most important things they need to know to assess whether a patient is a fit:
What area they are experiencing pain in
Whether they’ve previously had imaging done for the issue
Whether they’ve been told by a professional that they need surgery
Based on that, we proposed the following flow:
Unlike the intro screen for the mental health provider search, this one needed to include a bit more education on what orthopedic medicine covers. We hadn’t settled on what search terms would lead users into the flow, so I needed to write for people who had entered in “orthopedic doctor” (who we could assume know what that means) and “back pain” (who without some context setting might scratch their heads at being placed into this flow).
The diversity of users that would see this made the screen very difficult to write, and I started to get too tied up in technical definitions and the providers’ needs. Here’s one iteration of the screen before I figured out how to pare it back. I’d made it informative, sure, but I had also made it so long that no one was going to read it. I was really looking forward to getting input from real patients.
Research
We knew that a flow like this would be helpful to providers, but we didn’t know what patients would think about providing additional information about their needs, and whether they would see value in a flow like this.
As we prepared the prototype, I defined what I wanted to gain from our interviews:
Level of education needed
On the intro screen, does the copy help them understand why they’re there, and is the level of education on orthopedic care appropriate and appreciated?
Clarity on purpose, benefits, and our role
Is it clear that this flow is about narrowing their search rather than “diagnosing” what kind of doctor they need? It’s a subtle difference, but we needed to stay away from sounding like medical experts in any way or that we were collecting their medical information.
Here’s what we tested in moderated sessions with 8 participants (click or use arrow keys to advance, and R to return to start):
Users understood right away that this was about helping them narrow search results, and felt it would be worthwhile to continue.
The “Do you need surgery?” question yielded the most extensive and interesting feedback.
“For me I’d say no, I’m at the tail end of my healing.”
“Maybe if I’d been told by someone I’d say yes, but I may want a second opinion.”
“How would I know if I need surgery? That’s why I’m here.”
“I like ‘No’ and showing doctors that do other treatments so that surgery isn’t the only answer to the problem.”
People answered differently based on where they were in their care journey, and questioned “needing” surgery.
Research led us to make some significant changes: