Overview: Notifications are an essential part of the Facebook Experience. They alert users to different announcements and indicate actions users might take, and notifications help Creators stay in touch with their audience (Mugar, 2022).
Problem: However, Creators were frustrated with the volume, lack of relevance, and lack of organization of the notifications (Thompson, 2022). Rather than helping them accomplish their goals, notifications were getting in the way.
Solution: We designed notifications filters to help Creators identify and organize their notifications.
My role: UX Researcher
Duration: 4 weeks
Tools: Figma, Lookback
Methods: Design thinking, interviews, card sorting,
The purpose of this project was to:
- Understand how Creators use notifications
- Design strategies to help Creators manage their notifications
Ideation session
I conducted an ideation session on notifications with the development team, and generated filters as an idea for notifications management. We recruited in house Creators to come up with filter ideas, and we voted on the ideas to select the filter we would try first.
Mockups
I worked with the product designer (PD) to create mockups of the filter designs for the interviews. These filters were the result of previous studies with in-house creators that were conducted in the months before this study. The five types of filters were Comments, Mentions, Shares, Verified, and Currently Following.
Diary Study & Interviews
The goal of the diary study was to understand how Creators use notifications and to gather their feedback on the filters. Doing a diary study gave us insight into the day to day interactions Creators had with their notifications over the span of 2 weeks. Semi structured interviews allowed deeper investigation into those experiences, and also allowed me to engage in contextual inquiry by watching them interact with notifications during the interview as well.
Roadblocks
Internal recruitment was moving very slowly, despite a number of waves of email invitations. We wanted to match participants to online activity using internal data, but time was running short. With the potential added time, the team wanted to move ahead with the filter design without doing research.
Workaround
I quickly found an external panel of participants and incorporated questions about account type into a new screener.
I involved the team throughout the research process so they could integrate research findings in real time.
Data Analysis: I met with the team to watch videos together and discuss our observations, allowing for immediate action on findings. I integrated key highlights from the videos into the analysis.
Data analysis approach: Interviews
After I had the interviews transcribed, I integrated the notes from the videos then I coded inductively and deductively and used thematic analysis to identify patterns in the data. I analyzed survey data using descriptive statistics.
Insights:
Creators used notifications in ways we did not expect. They used them not only to moderate their audience, but also to monitor their page, compare performance of different content, and have insights into whether they monetize.
Next step: Codesign
We gathered the results from the study and ideas about possible filters. The team partnered with in-house group of Creators to do a codesign session and came up with new ideas for filters and partnered with the Creator Assistant (CA) team. CA is a way to quickly share insights with Creators when they first come into the app.
Impact: We came up with new ideas for monetization and unread notifications. We added new filter ideas from the codesign session including “currently following” and “monetized fans”.
We also gathered feedback on the filters, implemented new filter ideas, and started a collaboration with the Creator assistant team.
Meredith M Thompson 