Email: RE: Live Lockdown

Raw text

Transcribed text as recorded, including headers and formatting fragments.

Case 4:22-md-03047-YGR Document 2865-9 Filed 03/20/26 Page 1 of 7 AMENDED Exhibit 32 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 Case 4:22-md-03047-YGR Document 2865-9 Filed 03/20/26 METAMAAG-005-00069332 Page 2 of 7 (Meta) Zuckerberg Exhibit 47 3.27.25 mlr All Custodians Confidentiality Date Sent Metadata Cox_Chris;Jin_Kang-Xing; Sandberg_Sheryl;Zuckerberg_Mark; HIGHLY CONFIDENTIAL 2016/02/07 4:23 pm SEMANTIC SEMANTIC SEMANTIC Date Time Received From 2016/02/07 4:23 pm Mark Zuckerberg SEMANTIC SEMANTIC MDL BEGBATES MDL ENDBATES To Jav Parikh < ΜΕΤΑ3047MDL-014-00069332 META3047MDL-014-00069336 SEMANTIC SEMANTIC >; Maher Saba >; Sheryl Sandberg SEMANTIC exhibitsticker.com Case 4:22-md-03047-YGR Document 2865-9 Filed 03/20/26 Page 3 of 7 Message From: Sent: Mark Zuckerberg [/O=THEFACEBOOK/OU=FIRST ADMINISTRATIVE GROUP/CN=RECIPIENTS/CN= 2/8/2016 12:23:45 AM To: Jay Parikh ; Maher Saba ; Sheryl Sandberg CC: Chris Cox ; Fidji Simo ; Dan Rose ; Julie Zhuo ; Nick Grudin 1; Kang-Xing Jin [ ; Justin Osofsky ; Will Cathcart 1; Tom Alison ; David Wehner Subject: RE: Live Lockdown I meant to include him on the original note. + Wehner From: Jay Parikh Sent: Sunday, February 07, 2016 4:22 PM To: Maher Saba; sheryl Sandberg; Mark Zuckerberg Cc: Chris Cox; Fidji Simo; Dan Rose; Julie Zhuo; Nick Grudin; Kang-Xing Jin; Justin Osofsky; Will Cathcart; Tom Alison Subject: RE: Live Lockdown -david I'll loop in Wehner as well. Part of scaling live will require us to procure more capacity, which will be something we have to do in parallel to this lockdown. From: Maher Saba Sent: Sunday, February 07, 2016 4:15 PM To: Sheryl Sandberg; Mark Zuckerberg Cc: Chris Cox; Fidji Simo; Dan Rose; Julie Zhuo; Nick Grudin; Kang-Xing Jin; David Mortenson; Jay Parikh; Justin Osofsky; will Cathcart; Tom Alison Subject: Re: Live Lockdown Both Media and Infra teams are working on the Infra part. We'll come up with a plan this week. Saba From: Sheryl Sandberg <mailto Date: Sunday, February 7, 2016 at 4:04 PM To: Mark Zuckerberg < <mailto >> Cc: Chris Cox <mailto: >, Fidji Simo <mailto: >, Dan Rose <mailto >, Julie Zhuo <mailto: >>, Nick Grudin <mailto: +>, Kang- Xing Jin <mailto: +>, David Mortenson kmailto: >>, Jay Parikh <mailto: >, Justin Osofsky <mailto Maher Saba <mailto: Will Cathcart <mailto: >>, Tom Alison <mailto: Subject: Re: Live Lockdown Dan and the team and I will discuss incentive program and come back quickly Sent from my mobile On Feb 7, 2016, at 3:11 PM, Mark Zuckerberg <mailto >> wrote: Yup, I'm looking forward to discussing as soon as you have more details. Please use the framework I laid out below when thinking through the partner program and the tab design and functionality. On Feb 7, 2016, at 3:03 PM, Chris Cox < <mailto: >> wrote: We're been moving forward with pivoting the video and public figures teams in this direction over the last three days. We're all very excited about this direction. We'll share the full plan in the next few days. <mailto: >> wrote: On Sun, Feb 7, 2016 at 2:53 PM -0800, "Mark Zuckerberg" I'm excited about pushing ahead as fast as we can on Live because it is the most unique and promising new development in two spaces we care about deeply but are behind in: video and teens. I'm not suggesting Live will address all of our goals and problems in these two areas, but out of everything we're working on I believe that focusing on this product and executing it well is our greatest chance to make meaningful progress here. This can be the beachhead we need to expand into other use cases in video and teens. HIGHLY CONFIDENTIAL (COMPETITOR) HIGHLY CONFIDENTIAL ΜΕΤΑ3047MDL-014-00069332 METAMAAG-005-00069332 Case 4:22-md-03047-YGR Document 2865-9 Filed 03/20/26 Page 4 of 7 Our window to execute is limited though since we expect YouTube and Snapchat to launch their own Live products to much larger installed bases soon, and Periscope is already live in the market. We are the insurgent here and we must work quickly as one rather than acting like a slow incumbent. That is why I am proposing a lockdown sprint to launch something on ios and Android by the end of March. We had some debate this week about how much to emphasize Live in Video Home v1. We all have the goal of including episodic content in Video Home, but given our position of uniqueness and strength in Live compared to our relative weakness is premium content, I think it's critical that we start off with an experience that is focused almost exclusively on Live. Over time, we can add more premium and episodic content in as well. To make this more concrete, I spent some time composing how exactly I think this product should work for v1 and how it can evolve afterwards. This is more detailed and directive than I'd normally be, but since we're trying to mobilize quickly and need very clear specs, this seems like an appropriate way to push the conversation forward. At a high level, I see four major pieces that need to get built for a wider roll out: 1. Stable infra for scaling. We need to deprioritize whatever else it takes to get our infra able to handle a much higher load quickly. I have the least visibility into what is required for this and how far off we are, but this seems like it could be a very intense sprint. We should start this week. 2. Ramp up push notifs and NF ranking. If we want to grow reach, these seem like the two biggest levers we have. we should prioritize them to have this widely rolled out by the end of March. 3. Public figure incentive program. When we do our launch, we should have many public figures especially teens ready to start producing content. This requires some lead time, so we need to get started on this now. 4. Live Video / Video Home tab. We need a space people can go to find the best live videos. If we execute this in addition to all of the above, we will be the service with the best content, the most reach and the most focused experienced on live video (compared to YouTube or Snapchat). Points (1) and (2) are more technical and less product-focused, so I'm not going to spend more time elaborating on what needs to happen there. But they are obviously critical. We just need to be fully scaled and rolled out. I'll spend a little time focused on the public figure incentive program and then more time on the tab functionality and evolution. Public figure incentive program I think we should focus on having as much unique content as possible rather than paying for premium content that may exist on multiple platforms and that people are used to watching on YouTube. I also think we should be willing to invest significantly more capital than we've currently discussed to bootstrap our product for providing a destination to intentionally go to watch video. In the near term, I would spin up two efforts: First, a program that incentivizes public figures to produce Live content on Facebook. Live is inherently unique content unless you're streaming from multiple phones to different platforms simultaneously, which we can prohibit in our incentive program. I'd structure the program so that we pay people some minimum amount of money to produce some minimum amount of content - defined by number of Lives and overall minutes. That way they know they'll make enough money to be worth participating and they'll view it as part of their job regardless of whether the system is initially too small to be meaningful for them. On top of that, I'd also add a variable pay component that compensates these folks based on the number of total viewer minutes their content generates above some minimum. This incentivizes them to produce good content and to invest more once they see the model working. For the variable incentive, I think we could compensate these figures both for their Live and non-Live videos. I'd always make it so that we pay more for video exclusively on FB than content also shared on other platforms. This has a few nice properties. First, we're not paying people just for non-Live videos, so it doesn't set a bad precedent in the market since it will be clear the program is centered around Live and all participants will have contracts to produce a certain amount of Live content. Second, while focusing on Live, it also gives participants an incentive to experiment with and add their other content as they ramp up their Facebook presence. As the program evolves and bootstrapping becomes less important, I expect that we'd eventually eliminate the minimum guarantee part of the program and move to an all-incentive program. We can set that expectation upfront if you think it's required, but I'd also be fine committing to the minimum incentive program for at least a year or maybe two. HIGHLY CONFIDENTIAL (COMPETITOR) HIGHLY CONFIDENTIAL ΜΕΤΑ3047MDL-014-00069333 METAMAAG-005-00069333 Case 4:22-md-03047-YGR Document 2865-9 Filed 03/20/26 Page 5 of 7 and I think it would be reasonable to spend $100-200 million per year worldwide on a program like this to start. This would span teen celebs, YouTube/Snapchat/Vine stars, other already prominent figures ideally people who aren't strictly media entertainers, like athletes, journalists, scientists, doctors or even politicians if we can figure out a way to do that. It would be great to have this program rolled out with many dozens or even hundreds of partners for our end of March launch, so we should move to sort of the details of this within a week. Second, I'd invest in creating our own episodic short form content. Currently, a lot of investment goes into producing premium long form episodic content, but premium short form episodic is a new medium that barely exists aside from talk shows and use cases that can probably be well served by live. I recognize that if we start now, we won't have any of this content ready for 6 months or more. I am fine with that because I think we should focus on Live first anyway. But I think we should generally start working on this now so that by the time we want to have a supply of short form episodic content, this is ready. Eventually, if all goes well, this will be an even bigger part of our platform than pure Live content. This is an area I'd also be willing to invest ~$100 million per year in to start, although I'd encourage us not to go worldwide on this immediately. In addition to producing our own content, we'll also always pay for others' premium content through revenue share on ads. However, I think we should try very hard to not be too dependent on access to this content to bootstrap our products. Anytime we're paying more than our model supports for content others already have in order to incentivize consumers to use our product, we're playing a losing game. We should focus on getting unique content that fits our platform first and then once we've built a use case where people intentionally come to the tab to watch video, this other content will naturally flow to us as long as we have the right revenue share set up. So overall, my strategy would be to immediately roll out a broad incentive program to encourage many different kinds of people to produce Live and other unique content, while also ramping up a program to start producing our own short form episodic content. Live Video tab / Video Home Instead of discussing this in high terms like "emphasize live content more than other content", I'm going to offer a very specific spec for what I think this product should be. I'm of course open to discussing this and I'm sure this will evolve before shipping, but this is meant to be a specific articulation of what I think we should do. First, I'm going to discuss the different sections of the tab UI and what we're emphasizing. I'll also discuss how I expect this to evolve to add more episodic short form content over time. Then I'll get into specific features around notifications, badging and privacy I think we need to build and emphasize to make this work for our audiences. My proposal is that the tab has a single primary feed view made up of different sections. The UI can look similar to mocks you've shown recently for Video Home, but rather than having a single section, we will have multiple sections. There should be a big button to go Live at the top of the feed. The emphasis is on Live content over non-Live content, and within that on your subscriptions over other popular content. The sections are as follows: 1. Currently active Lives from your subscriptions 2. Most popular currently active Lives from across the network 3. Most popular currently active Lives from topics you're interested in (eg tech, politics) 4. Recent but not currently active Lives from your subscriptions 5. Recent non-Lives from your subscriptions 6. Infinite scroll of more popular Lives, recent Lives and eventually other videos I am assuming we will show ~20 pieces of content in each section, followed by a link to see more and then the next section. If there's no content in a specific section (eg none of your subscriptions are currently live, or you haven't specified any topics you're interested in), then we can just not show those sections. The default experience is that you show up and see the most interesting Lives currently going on. If there's something you've subscribed to then you'll want to see that first, and otherwise we should highlight the content most popular across the network and within your topics of interest. Since many people will not have any subscriptions Live right now or may not fill out interest topics, the default experience for many people will simply be a dashboard of the top 20+ Lives going on across the network right now. I'm purposefully suggesting we provide an experience very focused on Live at first. While YouTube and Snapchat can add in live features for their larger engaged video-watching communities, it will be HIGHLY CONFIDENTIAL (COMPETITOR) HIGHLY CONFIDENTIAL ΜΕΤΑ3047MDL-014-00069334 METAMAAG-005-00069334 Case 4:22-md-03047-YGR Document 2865-9 Filed 03/20/26 Page 6 of 7 difficult for them to have an experience so focused on live content. We'll have an advantage by providing a focused experience until we've developed a large community that has the specific intent to watch videos on Facebook that we can use as a beachhead to expand into more use cases. In the future, once this is more established, we can easily evolve the top section so that instead of only showing currently active Lives from your subscriptions, we also show recently active Lives and even other recent content from your subscriptions. At this point, the tab will probably just be made up of two algorithmic sections: your subscriptions and other recommendations. Eventually, I'd expect that we'd even merge these together and just have a single feed that blends together all of the different content people would want to see. I would not start here though, and I'd probably give this at least 6 months to develop before diluting the focus on Live at all. By the time we get around to doing this, we'll hopefully have developed some high quality unique content that can anchor the next stage of this experience. Returning to the Live-focused v1 I'm proposing, there are a few specific features that I think are probably critical to making this work. First, getting notifications, badges and the UI around subscriptions right is key for engagement. I'm assuming we'll notify you and badge the tab whenever someone you're subscribed to (and we think you'll be interested in) goes live. I'm also assuming we'll automatically subscribe you to your friends and everyone you follow, but will have to algorithmically prune some notifications to prevent people and especially pages from using this to spam. These notifications and badges will be critical for engagement. The notifications will let people know they can tune in now. The badges will last until you visit the video tab, so they'll also surface recent Lives that are no longer currently active as well. Since subscriptions are the trigger for sending notifications and badging, we need to make sure our UI really encourages and makes it easy to subscribe to everyone you want to hear from. Right now we're not doing that. We have the video overlay button which is persistent but subtle and does not describe what it does, and we have the subscription management screen you've proposed which feels like a power-user feature that you may not visit more than once. This Live-focused tab is more of a browse experience that will highlight lots of sources you're not yet subscribed to. Therefore, we should put much more emphasis on educating people on what the subscription button does, highlighting it and making it a fun interaction to use. Second, getting privacy right will be key especially for teens sharing tool rather than just for consumption. for encouraging use of Live as a Specifically, I would not emphasize friends as a privacy audience here. Instead, I'd emphasize public and building up a new sub-friends audience. The reality is, if a normal person broadcasts publicly, their friends will all get notified and that will be the majority of the audience they get. But they're explicitly deciding to share publicly, so they won't really mind if their parents, teachers or coworkers see them. When people really want to share with their friends though, I wonder if their "friends" audience is too big for content this raw. It feels like we'd be better off either encouraging them to make a new audience completely from scratch, or at a minimum automatically limiting it to people their age so it excludes parents, teachers, etc. I'm still undecided between rebuilding from scratch vs automatically limiting to people your age. Age- limiting is easier and may even be better over time since it's less prone to graph-bloat, but on the flip side it doesn't really help non-teens with context collapse and takes away some control. An ideal hybrid may be one where we present you with a checklist of people to either put them in or out of the audience, but where we start by hiding older people by default for younger people in our community. This would still keep it pretty simple and give people a good default while still giving full control. Finding a good privacy model is key to making this work for teens. If we don't do this, then even if the medium of live video sharing takes off, teens will eventually migrate to a different platform to do it. But on the plus side, if we build a new audience to enable people to do this here, then that audience could be useful for other forms of sharing across the service as well. Third, organizing into topics seems important for helping people form communities and find interesting content. Highlighting the top Lives across the network will be very interesting at first and will remain useful for discovering new and trending content. But once there is enough volume in the system to support sub- communities, people will quickly want to interact more around their own interests. For example, having Live be a platform where you can say you're interested in topics like football or tech or US politics or Pulis and then interact with other people around the world who are broadcasting about those topics would be magical and pretty easy to pull off. Some of this community behavior will be handled by people subscribing to the sources they find interesting. Some of it we'll be able to surface through algorithmic ranking over time. HIGHLY CONFIDENTIAL (COMPETITOR) HIGHLY CONFIDENTIAL ΜΕΤΑ3047MDL-014-00069335 METAMAAG-005-00069335 Case 4:22-md-03047-YGR Document 2865-9 Filed 03/20/26 Page 7 of 7 But in order to jumpstart this behavior, I think it would be useful to apply a basic tagging / taxonomy structure to as many Lives as we can. I'd take two approaches to this: manually tagging the most prolific Live creators to ensure they always show up in the right communities, and giving individuals the ability to tag their own broadcasts as well. When a person first uses the video tab, we could also ask them to tell us a few things they're interested in (which we'd guess at from their FB Likes), and we could use this to create sections of people who are currently live around these topics below the top section where we'd show the top current Lives around the world. This seems fairly simple to pull off, and very powerful for community building if we make it work. It would have a very strong network effort since rather than just creating a platform for broadcasting to lots of people, we'd also develop all these different sub-communities that would develop alongside the product as well. (As an aside, a more advanced version of this would allow people to filter based on not only topics but also gender, age and location. The interactive nature of Live makes it a great product for meeting new people in addition to just consuming content. It is inherently social. Of course, when you're interested in meeting people, it's helpful to filter them down to people you're actually interested in meeting the right gender, age and location. I think this would be incredibly powerful. There is also an element of just following attractive people. A lot of the most popular accounts on Instagram and Snapchat are models or attractive celebrities, and people just spend a lot of time watching their clips and looking at them. A lot of the Explore behavior on Instagram today is this. We could enable some of this through "fashion" or "celebrities" topics, but for the long tail of gawking I could definitely see people spending lots of time browsing public Lives of people of their preferred gender/age around the world. And it wouldn't be creepy because the people broadcasting explicitly want to have people watching them.) Ultimately, I don't think we need to build all of these filtering features for v1. I do think getting the notifications, badging and subscription dynamics and UI, the privacy model for teens and the most basic community architecture are important. Once we lay this groundwork, the community will tell us a lot of the features we need to build next. That's my basic summary of what I think we build. I didn't write much about making the infra work or ramping up notifications without spamming, so I just want to call those out one more time as absolutely necessary. If we had to choose between making those work perfectly and not building a tab at all, or compromising on those but having a good tab, I'd choose to make those perfect and not build the tab yet. I also would prioritize moving quickly and shipping whatever we can at the end of March over adding features. I think we should be able to do both if we prioritize well though! Let's get together to discuss early this week. In our next conversation, let's both surface any prioritization and tradeoff questions, as well as any questions and disagreements you may have over this general tab approach. I'm looking forward to sprinting on this and launching it by the end of March! HIGHLY CONFIDENTIAL (COMPETITOR) HIGHLY CONFIDENTIAL ΜΕΤΑ3047MDL-014-00069336 METAMAAG-005-00069336