Designing a personalised provider matching tool that gives top recommendations through personalisation and transparency
CHALLENGE
When searching for care, members assess a variety of factors in order to determine whether a provider is a “good fit” for them.
These include:
-
Expertise: Do they have the expertise + skills for my medical need?
-
Convenience: Are they conveniently located near my home or work (or virtual)?
-
Social proof: What’s their reputation? Do they have good bedside manner?
… and a long tail of other factors that vary based on the member and the type of care.
How do we create a matching tool that makes members feel heard, and we find the perfect provider for their needs?
DURATION
8 Months
IMMEDIATE TEAM
PM: Alex Haight
Eng: Megan Markle
Data: Catriona Smith, Zack
RESPONSIBILITIES
Strategy: Set the agenda and drive meetings, co-lead the project team alongside PM to define roadmap, and collaborate with cross-functional engineering and data science teams.
Research: Led the research initiative for the entire project, from discovery, surveys, moderated interviews to usability testing with 26 participants
Design: Led the user experience and visual design for the entire project, with a focus on creating an engaging interface that inspires trust in our members.
TOOLS
Research: Contextual Inquiry, Empathy Mapping, Service Blueprinting
Design: Figma
User Testing: Qualitative Interviews and Observation
Research Process
The research process involved discovery, qualitative interviews as well as usability testing with 26 representative participants, cumulatively spanning around 2 months.
My research process
Phases
Discovery Research
Deep dive into current behaviours around searching for a provider.
Focus: Understanding what are important qualities when finding a new provider and how they select and compare providers when presented with multiple options.
Contextual inquiry
1:1 moderated interviews with members
Focus: When looking at top recommendations and an explanation of why a provider is recommended, what builds trust?
Surveys / Unmoderated
Personalisation
Focus: Exploring what are attributes that members want to personalise on - what should be asked vs what should be inferred?
Usability testing
Validating final design solution using a interactive through prototype
Focus: Before heading into implementation, finalising the design flow by giving tasks.
Snapshot of a few participants
Participants
Representative members across different age groups, genders, ethnicities, income ranges, sexualities and socio-economic class
Key takeaway
Evaluating a provider
When engaging with a new provider, members often bucket their preferences within 4 categories. For primary care physicians, convenience and cost are given most importance while for specialists clinical quality and fit are given higher importance.
Getting down into the trenches
Understanding Member + Data needs
Based on my learnings from the research process, I created an exhaustive list of needs that are top of mind for members when they are searching for a new provider. We then started to correlate that to data that we currently have, data that would be easy to get, and data pieces that need more effort.
Design Process
The design process was heavily informed by research insights, and explored ways to make the interface feel more engaging and trustworthy.
Inspiration
Existing mental models around personalisation
Structure of the application
01.
Introduction to personalisation
02.
Intaking member preferences through a questionnaire
03.
Match results and top recommendations
04.
Provider details page
User flow
Module 1
Introduction
Introducing a personalisation questionnaire in an app can significantly enhance user experience by setting expectations regarding value prop, content and time spent.
Research Insights
Members already expect to be asked questions about their preferences, based on other digital experiences
Members are okay spending up to 5 minutes (and 7-10 questions) during personalisation
Design recommendations
Use language that make a member feel invested in finding the best match for themselves. Allow skip and return later if they want.
This is the hook - make the experience engaging using visual design and personalised illustrative elements.
Explorations
Introduction to personalisation
Final
Introduction to personalisation
Module 2
Personalization
Research Insights
Personalizing search results when looking for a new doctor allows individuals to tailor their choices based on a range of factors, ensuring a better match with their unique needs, preferences, and priorities. This, in turn, can contribute to a more positive and effective healthcare experience.
For PCPs, top attributes that members care about are Availability, Reviews, Distance, Network Status, & Gender.
Members are trying to personalize on their clinical needs as a priority, followed by race, cultural background, and language
Design recommendations
Focus on having questions & options based on the data we have but actively collect data for future roadmap items
Make the UI quick & easy so as not to cause fatigue, and play around with visual design to make it engaging.
Interaction exploration 1
"Dating app" style questionnaire
Exploring the interaction pattern of a left and right 'swipe' gesture - commonly found in dating apps such as Tinder. Playing on the idea of finding your 'right person'.
Interaction exploration 2
"Conversation" style questionnaire
Making it feel like you are talking to an actual person about what your ideal doctor would like it. Using the interaction pattern in fun games such as mad-libs.
Interaction exploration 3
"Wizard" style questionnaire
Using a more clinical approach - set of questions that look like a wizard or data app, helping you find the right doctor for your needs.
Final flow
Personalisation of care preferences
Final flow
Personalisation of provider attributes
Module 3
Match results
Research Insights
After users complete a personalisation questionnaire, it's important to present the results in a clear and user-friendly manner. Use positive language and visuals to celebrate the personalisation process - enhancing user satisfaction and the perceived value of the app.
Members want to see a consistent list of attributes on all cards as an easy way to compare, reflected from the questionnaire.
Members want to have controls over the results they see, and modify them if needed.
Design recommendations
Lean in to the top recommendations and add positive friction to see full provider directory.
Leverage visual design to make ‘Recommended’ provider cards feel special, while adding filters to customize others.
Exlporations
Top recommendations
Final flow
Top recommendations
Module 4
Provider details
Designing a provider details page involves creating a concise and visually appealing interface that provides users with relevant information about a healthcare provider. Prominently display contact information, locations and availability.
Research Insights
Members use a combination of reviews, recommendations, ratings, social proof, education & training to select a provider.
Members scour through multiple sources and sites to create a holistic perception of a provider before reaching out to them.
Design recommendations
For recommended providers, introduce a section to explain why the provider is a good match.
Introduce a section talking about provider quality, and breakdown of how it is calculated.
Exlporations
Provider details page
Exlporations
Explanation of quality
Final flow
Provider details page
Final flow
Explanation of quality
I would love to chat more regarding my in-depth process.
If you're interested, please contact me at tithi.jasani93@gmail.com or using the form below.