Modernizing Knowledge Panels

Google Search

 
 

Modernizing Knowledge Panels project focuses on improving accessibility to health-related information for all Google Search users by presenting it in easily understandable formats and creating a more visually engaging experience. As the Interaction Designer for this project, I played a key role in enhancing user interactions and ensuring a seamless and intuitive user experience.
This project was recognized by the Google Search VP in The Check Up with Google Health 2024

Goal

Our objective was to optimize the presentation of health-related information on Google to meet the diverse needs of users, ultimately empowering them to make informed decisions about their health.

Approach

To achieve this goal, we aligned with the new Search framework to ensure consistency throughout the Search Results Page (SRP). Additionally, we evaluated and modernized the existing Health Knowledge Panels (KPs), addressing inconsistencies, outdated elements, and lack of cohesion across the user journey.

Background

A 2019 study revealed that 89% of patients in the US search for health symptoms and conditions on Google before consulting a doctor. This underscores the importance of providing accurate and accessible information to users seeking health-related guidance online.

The KP UIs have not been updated in many years and carry over multiple legacy design frameworks. They are inconsistent, dated, hard to scan, and lack cohesion across a user journey.

Discovery & Idea Validation

One of the artifacts I created to gain insight into product needs and user pain points was a UX Audit of the current Knowledge Panels and symptom search user flows. Through this audit, I identified that the KPs utilize multiple legacy design frameworks and display inconsistencies across various symptom and condition queries.

Concept Testing

Overview

To implement the new framework for health-related searches, we conducted concept testing to understand how users prefer to interact with various types of information for symptom-related health queries and assess the efficiency of different proposed visualizations and navigation methods.

I collaborated with a Researcher, Product Designer, and Project Manager to develop 3 different prototypes for concept validation with real users. These prototypes were created using Figma.

Objectives

  • Comparing different treatments for symptoms within the new framework and gathering feedback on their effectiveness.

  • Understanding users' evaluation of available information and their preferences for information sources.

  • Identifying the most useful types of information to include in the framework.

  • Gathering ideas for future improvements.

Three Prototypes have been used to validate the concept with users.

For confidentiality reasons, I have omitted the actual prototypes that underwent testing.

Results and insights 

  • Most participants search to learn about symptoms and identify possible treatments before deciding whether to self-treat or visit their doctor

  • Most participants preferred the Inline Expansion design concept

  • The highlighted content inside the boxes was appreciated and the preferred expanding boxes to show the additional information instead of taking them to another page

  • Most participants valued knowing the source of health-related information and expected to see the source in proximity to the content

Design Phase

Collaboration with framework leads has been established to utilize modified versions of existing components, tailored to meet the specific needs of health content.

Strategic Solutions

  • Dynamic ranking

    Implement a robust dynamic ranking system that integrates Google-curated content (in partnership with Mayo Clinic) and third-party content.

  • Inline expansion

    Highlight the most relevant content by leveraging the Inline Expansion framework.

  • Visual experience

    Create a visually enriched health experience by utilizing Universal Images.

  • Attribution

    Attribute health content clearly to indicate the source of each piece of information.

Four unique strategies were tested in a live environment, each showcasing content with different presentation methods.

For confidentiality reasons, I have omitted the actual high-fidelity mockups that were tested.

Implementation Phase

Work with engineers to ensure product quality standards by meeting regularly with the team and conducting a bug-bashing process.

Product Release

Option number 4 demonstrated better performance in the live experiment, achieving approximately a 50% improvement in content accessibility and enhancing users' ability to find the relevant information they needed.

For confidentiality reasons, I have omitted the actual values for these metrics.

Next

Cancer Care - Kaiser Permanente