Find a New Doctor with Efficiency and Joy

Competitive analysisProduct strategy alignmentWireframingVisual design

I helped design Healthgrades' guided search, making it easier for users to find and book the right healthcare provider quickly. Through user-focused design and testing, we created an intuitive and conversational experience.

The Quick Take

TLDR: I led the design of Healthgrades' guided search feature, helping users efficiently find and book appointments with local healthcare providers. Collaborating closely with development, we delivered the conversational, non-linear search interface in just 4 weeks. User testing confirmed high engagement, with 80% of participants favoring the new approach over traditional linear questionnaires.

My Approach

Background

Healthgrades.com enables over 1 million users a day to research and book appointments with local healthcare providers. Most of the time, people were well into their search process for a doctor by the time they visited Healthgrades, often seeking to verify a recommendation. In an effort to showcase the power of our physician data and patient feedback, the executive team proposed I build a guided search experience. This feature would either validate the original search or suggest better matches, ensuring users find the ideal healthcare provider quickly.

Discovery Research

After defining the project goals with my product owner, I dove into a thorough market analysis to see what currently existed for step by step questionnaires out in the world wide web.

Oscar's animated style and bold colors were great, but the tone felt robotic.

In addition to the bloated questionnaire, WebMD included these very terrifying avatars of the body. 

What I was looking for was something more conversational. The mattress company Beauty Rest made me feel like there was a human on the other side of the screen by their use of illustrations and UX writing and even though they had 6 pages of questions, it didn’t feel burdensome to me. This was the experience that most closely aligned to what I wanted to achieve.

Her facial expression changed as you adjusted the slider, a memorable moment of delight.

I also joined multiple dating sites in the name of research and even though it’s been a few years since completing this project, JDate’s parting of the Red Sea loading state still stands out to me as one of highlights.

Whiteboard Extravaganza

Initially, I was tasked with building a step-by-step wizard experience. However, both the product owner and I shared the hypothesis that users might prefer a more natural, non-linear flow of questions to receive the best recommendations. But envisioning this concept was still fuzzy to me.

To tackle it, I decided to clear our calendars, grab a sleeping bag, and book a spacious whiteboard room for two days. This gave us the freedom to brainstorm and build upon each other's ideas without constraints.

We started by listing all the data points we had on our providers. Then came the pivotal question: How could we ask users only the questions they truly wanted to answer? This was the crux of the issue with the linear guide. 

As we neared a defined solution, I would dash upstairs, tap our data science teammate on the shoulder for their input, and then update our designs based on their technical feedback.

Ultimately, my solution was to present all the filters on a single screen. Based on the user's selections, we would then prompt follow-up questions. In this flow, the user gets to prioritize what matters most to them.

Our final whiteboard concept ready to be digitized.

Testing the Design 

To test our hypothesis, I created a low-fidelity prototype for each search version. I called the conversational/natural version 'Salt' and the linear version 'Pepa'.

Then, I conducted two series of unmoderated tests using Usertesting.com. In the first round, participants navigated through both search experiences, and the results largely favored Salt. However, I was concerned that the test might be skewed since I asked users to find doctors of different specialties for each experience.

In the second round of testing, users were given one set of criteria to filter both experiences for an apples-to-apples comparison, and the results remained consistent, with Salt being the preference.

A reference to the acclaimed girl hip-hop group of the 90s.

Salt Prototype

Pepa Prototype

Visual Polish

Entering the high-fidelity phase, I was given approval to utilize our new brand elements. I designed the experience as a full-page modal, maximizing white space for a clean look. Inspired by Beauty Rest, I then created a playful set of illustrations to represent each set of questions. A pale cerulean circle was used to frame each icon, and then a subtle background pattern was added for texture.

A high-fidelity prototype of Salt for development handoff.

Rapid Design-to-Development

After delivering the Salt designs to my developer teammates, we collaborated to push them into production in less than 4 weeks for an A/B test. The results were promising: an 81% question completion rate and a 45% conversion rate to provider profiles.