1. Case Study: Messenger

With the sheer amount of time that we have dedicated to using our mobile devices, day-to-day annoyances begin to crop up constantly. Starting with informal recognitions of negative experiences using certain mobile apps, I decided to target one that had provoked significant critique and friction—Facebook’s Messenger. I use Messenger as my primary chat platform to communicate with friends and family, like many others, as it exists as the world's second-most popular platform after Facebook’s own Whatsapp. Popularity, however, hardly indicates an ease-of-use.

Due to frustrations with successive Messenger updates that continued to complicate the application, I have had ideas spinning in my head about improving this experience for over a year. Thankfully, with data, surveys, or their own realizations, Facebook finally released Messenger 4 for all of their users—their latest update that attempted to clean up this increasing clutter (see: "Decluttering the home screen...from nine tabs to three").

My frustrations were mirrored in their own study, where they noted that 71% of their surveyed users believed that simplicity was a top priority for a messaging platform. However, I still believe that there continue to be facets of the application that could be pushed to further enhance a user’s experience.

For one, the majority of my frustrations stemmed from the antiquated system of managing chat threads—an issue that does not only effect Facebook, but throughout the entire industry. I found the system of having the most recent message first unhelpful for my usage of the application, as it often gave snippets of text that completely lacked information or context. The lack of instantly available information, as well as my tendency to forget about each topic for each individual thread, compounded my frustration.

In order to tackle my primary issue of an unhelpful home screen experience, I began thinking about contextual summaries about your message threads. With Messenger's assistant, M, I realized that Facebook already had many of the technologies that could allow this to be possible. However, I still needed to gauge interest and frustration to validate my claims.

With the survey, I asked for the amount of chats that users were engaged with primarily to confirm my previous assumption that the amount of chats that was immediately available to the older Messenger 3 was excessive. Messenger 4 better highlights each individual thread, which I agree with. A larger number of threads that an individual participates in also means a higher probability to juggle more topics, which I highlighted as a problem.

With 53% of respondents finding current chat previews as not helpful (neutral to highly un-useful), the study highlighted the fact that the home page was underutilized. During in-person interviews, though some users commented on the helpfulness of confirmations and simple replies with a glance of the home page, the majority of users preferred to click into the chats to figure out context.

Studies from the Center for the Governance of AI indicated that 67% of respondents had little or no confidence in Facebook for developing AI, mirroring my own findings of 75% of respondents with a lack of trust for FB and their use of AI. With a similar unlikelihood of using Messenger's AI assistant amongst my surveyees, I understood that concepting an AI product would mean addressing the needs of the privacy-conscious.

With my survey results, I noted four additional issues:
1. Button Placement and Iconography
2. Artificial Intelligence as Untrustworthy
3. Privacy Issues
4. Lost Place In Chats

Within the scope of my user experience project, I could not address 3. Privacy Issues on Facebook's data collection end. I could, however, personally address the potential for users to turn off my primary AI function to create contextual summaries, and thus, targeting the concept of 2. Artificial Intelligence as Untrustworthy.

The contextual home page itself addressed my primary frustrations, but if a user were to not utilize the M functionality, I needed a suitable replacement.

I came up with the initial idea of pinnable notes for yourself or the entire group. It would address 4. Lost Place In Chats for those of whom did not use the contextual AI, but also help users browsing their home pages who did use the service, as it would remove the potential for mistakes made by the AI.

With both user segments having a suitable replacement to the current system, I realized that I could concurrently resolve an issue with 1. Button Placement and Iconography, where many users accidentally hit the video and audio chat buttons in the top right corner. In the current system, these accidental button presses cause significant negative reactions, as large cards indicate that a call had been made, and thus often followed with the user apologizing about the mis-press.

The Pinned chat extension could alleviate these pain points by creating additional friction to the video and audio chat functions by making it more difficult to access them, as well as bring further functionality to Messenger without needing to extensively teach the user new actions.

With user complaints with losing one's place in a conversation and my desire to address 4. Lost Place In Chats, I believe that the Pinned chat extension alleviates some of those user pain points. Not only are users able to jump back into their chats knowing the general context of their chat, previously pinned messages could send users back up the thread. Overall, these additions give users more control over previous messages and their information.

This concept exists within an interesting space: one which embraces AI for better user experiences, which is often the direction that tech industry heads towards; but another that understands privacy-conscious users. Like many other design objectives, being able to target the needs of a specific group often results in benefits for the larger whole. That also isn't to say that this is the last iteration of messaging—it is, however, a look into the possible futures of messaging. How are companies already processing our data with AI, and how can they make it more accessible to the regular consumer?

Direction, Design: Reginald Lin
Made with Figma, Illustrator, Photoshop