Email: we'll need to be very good about not notifying parents / teachers

Raw text

Transcribed text as recorded, including headers and formatting fragments.

Case 4:22-md-03047-YGR Document 2757-13 Filed 02/20/26 Page 1 of 3 AMENDED Exhibit 205 PLAINTIFFS' OMNIBUS OPPOSITION TO DEFENDANTS' MOTIONS FOR SUMMARY JUDGMENT Case No.: 4:22-md-03047-YGR MDL No. 3047 In Re: Social Media Adolescent Addiction/Personal Injury Products Liability Litigation Message Case 4:22-md-03047-YGR Document 2757-13 Filed 02/20/26 Page 2 of 30ctober 24, 2024 EXHIBIT Meta-Jin-18 From: Sent: To: Subject: [/O=THEFACEBOOK/OU=EXCHANGE ADMINISTRATIVE GROUP (FYDIBOHF23SPDLT)/CN=RECIPIENTS/CN=MAHERSABA541] 2/29/2016 11:38:13 PM Kang-Xing Jin [kxjin@fb.com] FW: Live, teens and graph issues Hi KX, should be on this email? On 2/28/16, 11:21 AM, "Mark Zuckerberg" <zuck@fb.com> wrote: >of all the efforts we're working on for teens, I'm most optimistic about Live. This is somewhat because of the exclusive content we'll get for consumption, but the much more powerful reason is because it's a new form of raw sharing. > >However, I'm worried that even if Live is a new raw format that young people enjoy, they may quickly migrate to a standalone product with a clean graph if we don't fix our graph issues for this audience. That is, they may like Live, but still not want to live stream to their parents and all their FB friends. > >It's not hard to imagine a world where we are first to scale Live and get people excited about this, and then a standalone product or one with a cleaner graph like Periscope or Snapchat is the major beneficiary of this market and us showing the world how awesome this is. > >We've seen over and over that with social use cases adjacent to our current ones, the key is to bootstrap using our distribution and graph but to also allow a unique graph to grow independently and not keep it too closely tied to our current one or it will stifle the new ecosystem's growth. > >We saw this with messaging, where our distribution and graph helped bootstrap it but being too closely tied prevented it from addressing the full market. We are now working on disassociating the graphs somewhat to grow to a global maximum instead our current local maximum. We also see this dynamic playing out with Marketplace now where our distribution and graph are helpful but some independence is also helpful. In each of these cases, our rigidity in sticking with our existing graph created space for competitors to surpass us with cleaner and more flexible graphs. > >I'm pretty confident this dynamic will play out in any social use case around live video for teens where our graph is already an issue. > especially >Therefore, we need to make sure we simultaneously use our assets to bootstrap while also building the right graph structure to allow the ecosystem to reach its full potential. > >Right now we're focused on shipping Live and getting the right content. I also think we need to think about how to make the graph work especially for young people. > >Here are some specific questions, thoughts and ideas although what I really care about is that you guys focus on this and take this into account in what we ship. > >First and most important, for social (friend sharing) use cases, how do we choose which friends to notify that you're live now? > >If we tell teens' parents and teachers about their live videos, that will probably ruin the product from the start. At the same time, even blasting notifs to many of your young friends may require a different graph to make sure you don't feel like you're spamming people by going live. My guess is we'll need to be very good about not notifying parents / teachers, and then also use machine learning to find the optimal balance between sending enough notifs to get producers enough audience while also not having people feel like we're spamming. Since the brand effects of feeling like live video is spammy will be long term pain that's harder to measure in the short term compared to the short term benefits of getting people more audience, we should probably be a bit more conservative in the near term and send somewhat fewer notifs than the short term metrics indicate. > >Next, for privacy/audience, should we default to existing FB friends, ask people to curate a sub-list of their FB friends (perhaps removing parents) or ask people to specify a new audience entirely? > >I think it's pretty important that we bake into the product early on the expectation that this graph will deviate from your existing FB graph, so I'd be inclined to not just blindly default this to using the default FB friends list. I'd try to make editing the live video graph a reasonably prominent part of the experience, perhaps by having a persistent icon for managing this graph on video home next to the button to go live. We may also want to prompt or encourage people to edit this when they go to video home, but I probably wouldn't roadblock them before they go live or insert any friction into the experience of going live from the NF composer. HIGHLY CONFIDENTIAL (COMPETITOR) HIGHLY CONFIDENTIAL – ATTORNEYS' EYES ONLY ΜΕΤΑ3047MDL-020-00476655 METATNAG-005-00606807 Case 4:22-md-03047-YGR Document 2757-13 Filed 02/20/26 Page 3 of 3 > >It's probably better to get people to edit their existing graph (perhaps by suggesting people to remove) than asking them to build a new one from scratch. For what it's worth, I think Snapchat's UI for controlling who can see your story is pretty good it's just a list of your friends with a checkbox next to each of them. In the context of our graph where you are likely to already be friends with your parents, making it easier to uncheck them or other sensitive people specifically would be useful. > >Notifs and privacy cover the two most extreme aspects of the graph who we aggressively ping about your activity and who can see it at all. There are of course intermediate levels of notification that we should consider as well, like which friends badge your video home tab and how we rank different live content in News Feed. My sense is we'll need to find a balance where we are very aggressive about showing live content from people you care about, but we need to make sure we're much less aggressive with people you may not care about. The more disruptive the channel, the more we need to make sure we don't send you lives from people you care less about. So this is most critical for notifs, a somewhat wider funnel for badging and somewhat even wider for News Feed. But even for News Feed we should make sure we're not up ranking live content from people you don't care about just because it's live and we think live content is good overall. > >A related issue with sharing live content is ephemerality. I've done a few lives so far and sometimes when they don't go well I think about whether I should delete them afterwards. It's not that they were a bad experience, and I've always felt fine about the people watching live seeing them, but I just don't always feel like I want the video to stick around forever. This isn't strictly a graph issue but it gets into the social dynamics around how people want to share live content. Of course people have the ability to delete a live post afterwards the same way they'd delete any other post, but right now our UI doesn't signal that this is a suggested or recommended behavior. It may emerge naturally like deleting photos has on Instagram, but it may be healthier to make this a part of the actual UI. That way people can feel like they have two options: livestream+record or livestream only. I'm not sure exactly how I'd design this UI. I don't think we should ask people to decide upfront whether they want to save it since their decision will likely depend on how the live goes. But we may want to give people a choice at the end in some way of whether they want to hide it. We'd need to make sure this isn't too aggressive and doesn't encourage people to hide it when they would otherwise be fine with it, so this is a tricky balance. But finding some way to acknowledge this behavior would probably be healthy. > >These are the many social and graph dynamics I've thought about so far. There are probably others. My main goal is to raise that this is important especially for young people but likely for everyone if this is going to be a new form of sharing. We should find ways to build this into the product from early on. HIGHLY CONFIDENTIAL (COMPETITOR) HIGHLY CONFIDENTIAL – ATTORNEYS' EYES ONLY META3047MDL-020-00476656 METATNAG-005-00606808