Find My Fit: A Messaging App for Healthcare

Recently there has been a trend towards a messaging-based economy. Messaging platforms like Facebook Messenger and Snapchat provide interesting ways for businesses to interact with users. One of the more interesting ones is virtual assistants (or bots) that use natural language to communicate rather than user interfaces. Take a look at Chris Hawkins’ blog to learn more about my team’s perspective on virtual assistants.

With the fragmented healthcare system and the complexity of medical information, consumers can be overwhelmed by the multiple and nebulous sources of information, hindering their search for care services. My team at Accenture set out to explore how an insurance company could provide a personalized user interaction in the healthcare space using a natural user interface with the “Find My Fit” messaging app.

The Current Online Experience

Quite often health insurance sites are perceived to be cumbersome and repetitive by users. Even though the insurance company has an abundance of patient information at its disposal, its site requires users to fill out a long form and select numerous filters. Some data input should be unnecessary once the patient has logged in with a secure password. An example of an insurance website is shown below.

An insurance website (name changed)

An insurance site is not necessarily the first site that patients will visit for health issues. Patients look at multiple sites to research and gather health information. Some well-known examples of related activities are:

  • Inspiration and Aspiration: To set and reach their health goals, people share nutrition and exercise routines (e.g. Pinterest).
  • Recommendations: To learn from others who have gone through similar experiences (e.g. Yelp).
  • Advice: To check symptoms or get second opinions online (e.g. WebMD).
  • One-On-One Attention: For personalized care at any location (e.g. Teladoc).
  • Recording Your Quantified Self: To track data, such as weight or miles ran (e.g. Withings).

Pinterest, Yelp, WebMD, Teledoc, and Withings are all companies that have changed consumers’ digital expectations.

Find My Fit

The Find My Fit experience is designed to be intuitive and dynamic. We’ve implemented a conversational UI tailored to the individual based on their preferences and activities. Scheduling an appointment can be streamlined to a quick conversation by automatically using information from a patient’s calendar, appointment history, and insurance data.

Guiding Principles on Data

Accurate data is paramount to personalization when designing message content. Our solution consisted of three parts:

  1. Data Collection: Data is collected at every stage. CRM, appointment meta-data, social media, direct user input, and clinical data contribute to a complete picture.
  2. Feedback: Requests are made to solicit feedback to improve app predictions and recommendations.
  3. Data Confirmation: Users are asked to confirm the output of data analytics and verified data is given higher confidence

Surveys are triggered using GPS and time to request feedback at the right time and place. For example, a user can give immediate feedback when they arrive home after an appointment. The feedback automatically adjusts future doctor matches.

App Design

The inspiration for the design of the messaging UI came from the Quartz news app. Like Quartz, the Find My Fit user is given a few response options, which are short and conversational.

A Find My Fit Messenger page interaction

As shown in the video, the app uses basic information when the user opens the app for the first time, including the dependent’s name and user’s browsing history. Over time, the conversation will become more personalized as the user provides additional information, such as preferred mode of transportation and preferred appointment times.

Rich inputs were used for whatever text couldn’t do. To provide a few examples, we included social media login pages so users can get improved recommendations (e.g. nearby fitness events), map views to choose a clinic location, carousels of doctor choices, and star ratings for feedback.

Three types of rich inputs

We also thought about how a user could initiate a new line of questioning on the Messaging page. This menu gives the user an opportunity to find new insurance, find a doctor, chat directly with a doctor, and schedule an appointment with a doctor.

The second tab is the Profile page, which lists a user’s health teams. If Lucy had dependents in her insurance plan, the names would be listed as part of “Your Team”. The insurance information is provided when the user selects a name. “Medical Team” is updated immediately after selecting a doctor on the messaging page. Lucy can access Dr. Jenkins’ patient reviews, clinic location, medical education, and more. Though it’s not included here, there could also be a “Home Team” for caretakers or a “Fitness Team” for physical therapists or fitness instructors.

The third tab is the History page, which allows a user to review important events without scrolling through all the messages. Events like a doctor appointment or a prescription refill reminder could be listed as well.

Interested in hearing more?

Please contact me or my team members (Chau Dang, Chris Hawkins, or Dave Nguyen). To read more about Accenture Technology Labs and our R&D areas, visit our website.