MyShake
The Berkeley Seismology Lab's earthquake early warning network powered by citizen scientists
Duration
May 2017 to Oct 2018
My Role
Designer + Researcher
Team
3 Developers, 1  Lead
MyShake is a citizen science app from UC Berkeley that uses smartphone sensors to detect earthquakes. With enough users near an earthquake’s epicenter, MyShake can send early earthquake warning (EEW) push notifications. I was the project's first UX researcher and designer, aiming to make MyShake a robust earthquake resource and attract new users to create a more reliable warning network.

Constraints

Outcome

MyShake's redesign was released in Summer 2019. In October 2019, MyShake was selected by the U.S. Geological Survey to become the first mobile app to deliver ShakeAlert earthquake early warnings in California.

Baseline Research

Who uses MyShake and how is it used?

Since MyShake was looking to expand their user base, I worked to establish a product baseline and identify opportunities for growth by deploying a survey to users, conducting usability testing, and analyzing Google Play analytics.

Insight 1: Unintentionally narrow user base

Of the 208 users who responded to the survey, most were above 35 and were highly educated which didn't align with MyShake's goal of serving all types of people in earthquake prone regions.

90%

of respondents were over the age of 35

63%

had a bachelors degree or higher

Insight 2: Organic acquisition, low retention

Between 2016 and mid-2017, MyShake had 296,000 downloads. Its mission drew media attention and organically attracted new users but low retention rates show that most did not become long term citizen scientists.

Insight 3: Friction and unmet needs

User testing and feedback revealed that the functionality of the app was limited and many earthquake-related needs were not addressed. The existing UI also introduced friction in key areas.

Insight 4: Increased usage post-earthquake

It wasn't all bad news. While we identified several obstacles, we also noticed an interesting trend: downloads and usage spiked in the hours and days after earthquakes.

User Research

How do people respond to earthquakes and what challenges do they face?

To develop a reliable early warning system, MyShake needed to grow beyond its niche user base.  Analytics showed us that directly after earthquakes downloads and usage increased significantly. Our team decided to leverage this trend and work to become a more robust earthquake response tool. I researched earthquake response behaviors and unmet needs to understand how MyShake could both serve and retain more users.

Literature Review and Expert Interviews

Memories of earthquakes are often distorted by stress, making standard user interviews an incomplete tool in learning about earthquake experiences. I adapted by reviewing literature and interviewing eight experts in the fields of earthquake outreach and emergency behavioral responses, synthesizing my findings in a task model.

User Interviews

My team also interviewed more than 80 Bay Area residents to learn about the mindsets of those living in a seismically active area. I recruited four of my team members to help conduct these interviews and their participation was critical to reaching more participants. It also helped generate team buy-in about the importance of user needs later on.

Insights

We found that fear associated with earthquakes often led to avoidance and a sense of helplessness. However, we also identified several themes that could help us empower users.
CONFirmAtion

After earthquakes, people sought avenues to share and confirm their experience.

IMPact Assessment

Before responding, people looked for information about an earthquake's impact.

Acknowledgement

Citizen scientists need contributions to be acknowledged  to stay engaged.

Design Response

How might we empower people to respond after an earthquake?

Our baseline research revealed that MyShake usage spiked in the hours post-earthquake and this was supported by the sentiments towards earthquakes expressed during our user research. This trend and the many needs we identified in the space led our team to focus on the problem area of immediate earthquake response.
I created user journeys based on our research and sketched concepts to identify opportunities and explore ideas with my team.
We narrowed down ideas based on user impact and technical feasibility, deciding to move forward with several new features designed to empower people in the minutes and hours after experiencing an earthquake. By demonstrating value as a post-earthquake resource, we aimed to grow MyShake's user base and create a reliable earthquake early warning network.

Experience Reports

Research Insight
People seek avenues to share and confirm their experience after an earthquake.
Response
Provide an experience survey for each earthquake where users can report shaking intensity and infrastructure damage.
Considerations
User’s time and attention are limited, especially after an earthquake, and traditional experience report surveys can be cumbersome and confusing.

Experience Report Impact Map

Research Insight
To determine the appropriate response, people need information about earthquake impact.
Response
Share "Experience Report" responses, using an interactive map and data visualizations so that users can asses the overall impact of an earthquake.
Considerations
In emergency situations, information needs to be clear and digestible at a glance. It is also critical to ensure the validity of reports and protect user privacy.

Contribution Dashboard

Research Insights
People become less engaged during periods between earthquakes and need acknowledgement of their contribution to stay engaged.
Response
Create a dashboard where users can track their participation as citizen scientists and the project's overall progress.
Considerations
Contributions should be framed as metrics that grow and change over time so users stay engaged.
Using the lens of immediate earthquake response, I began reimagining MyShake's architecture, flow, and existing features to improve current pain points as well as align them with new problem space and features we’d identified.

Earthquake Map and Details Page

Existing Pain Points
Events were hard to find as they were color coded based on date rather than magnitude and were represented by small, hard to select circles. Tapping on an earthquake led users to an external webpage created for experts rather than everyday users.
Response
Earthquake icons colored based on magnitude make it easy to understand nearby activity at a glance. Increasing the icon size as well as providing a list view and keyword search makes events easier to find. When users select an earthquake, they are taken to the in-app visualization of experience reports.

Notifications

Existing Pain Points
Users were limited to setting one notification type, despite needing more robust monitoring options. It is important to note this feature is not an early warning, rather a notification after an event has produced shaking.
Response
I redesigned the notifications settings to reflect users needs, including awareness of earthquakes in their immediate area as well other areas of interest and alerts about significant global events. Switching from radio buttons to a sliding selector also allowed for more precise customization.

Iteration

How do users respond to the redesign?

After outlining new features and design updates, we began refining our designs through usability sessions. Our first round of usability analysis was focused on an Invision prototype. Our second tested a functioning front-end interface with placeholder data.

Experience Reports

  • Provide (many) escape routes. Users initially felt trapped in the report flow. We added more exit routes, the means to skip and return to questions, and a mechanism to store incomplete reports for users to finish later.
  • Make choices easy. Reports took longer to complete when users vacillated between options: "What's the difference between light and medium shaking anyway?" We added imagery and descriptions of each shaking and damage level to support their decision-making.

V1

v2

v3

Experience Report Impact Map

  • Make data visualizations clear at a glance. Users found clustered visualizations of reports too specific. They were more interested in the impact to overall areas rather than specific points. Using a bin grid to display reports made the impact easier to digest at a glance.
  • Base micro-copy in users' language. Many users associated the "Share your experience" button with sharing to social media. A more specific call to language helped align users' expectations with the intended action.

V1

v2

v3

Design System

Intentional + Lightweight

I created a design system for MyShake that is simple, concise and supports the goal of bringing consistency to the interface while being easily maintained by a small, fast-moving team.

Outcome

Share and contribute

By answering just three questions users can quickly share their experience, helping inform their community and contributing to research.

Understand impacts

An interactive map and data visualizations help users understand how an earthquake has affected their community.

Track your contributions.

Users can see the value of their contributions with a dashboard displaying personal and project metrics.

Reflections

Speak your stakeholder's language. The first hurdle of this project was convincing project leaders to invest in human-centered design. The key decision-makers are seismologists driven by MyShake's research aims. I used low cost methods like app analytics analysis and man-on-the-street interviews to establish data-based findings that tied user needs to their research goals, effectively convincing them to invest in serving users.
Synthesize solutions from diverse perspectives. Bringing in the voices of a wide array of users and experts helped us understand the spectrum of perspectives in our larger target user group. Similarly the diverse perspectives on our team, ranging from project managers to seismologists, led us to think more creatively and develop well-rounded solutions.