Meta & Children's Online SafetyEmail: feedback requested: unship proposal for Time Spent
Raw text
Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 1 of 15 AMENDED Exhibit 260 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 2889-42 Filed 03/27/26 Page 2 of 15 Message From: [/O=THEFACEBOOK/OU=EXCHANGE ADMINISTRATIVE GROUP (FYDIBOHF23SPDLT)/CN=RECIPIENTS/CN= Sent: To: Subject: 8/19/2020 10:53:43 PM ; Liza Crenshaw Re: feedback requested: unship proposal for Time Spent (in "Your Activity") Attachments: image001.png; image002.png Separate from the unship discussion, I think these quotes are worth raising visibility on. Seems worth a group post (vs. email) for visibility. From: Date: Wednesday, August 19, 2020 at 9:02 AM To: Liza Crenshaw Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") Hi! I know this conversation is currently paused, but wanted to share another reason why it's really crucial to keep (or update) Time Spent tools. Much of the coverage around the Feed Recs launch has been around IG caring less about digital well-being, but did note that IG still has tools to address time spent. Important for us to have proof points to counter negative press around well-being (knowing that Feed Recs is a priority for IG). The Verge: "The move feels like a major reversal from Instagram's stance two years ago on screen time. When it launched the end of feed notification, the company's former CEO and co-founder, Kevin Systrom, positioned it as a decision made with users' mental health in mind. He said at the time that the company was working on "tools that will help the Instagram community know more about the time they spent on Instagram," adding that "any time should be positive and intentional." But suggested posts could easily keep people on the app for longer." Fast Company: It's hard to see how this feature is not another play to feed Instagram addiction, but the company has put guardrails in place: Once you scroll to the end of your feed, a message pops up asking if you are interested in seeing more content that might appeal to you from accounts you don't follow, and you must consent before it loads.... Like everything on Instagram, nudging users to spend more time on the app so Instagram can serve them more ads is at least part of the endgame. From: Date: Friday, July 24, 2020 at 1:09 PM To: Liza Crenshaw Subject: FW: feedback requested: unship proposal for Time Spent (in "Your Activity") Hi / Liza. The time spent unship proposal went through the IG Unship Review process (w/ Checking: and came back as unship. Is the level of concern from Outreach expressed by you two and Karina the same now, less or greater, given current climate and our recent decision to pause Mental Health as a priority for H2? HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181774 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 3 of 15 A next step here is to gather new information (ex. current Outreach assessment, recent MH decision) and follow up (and maybe escalate). From: Date: Friday, July 24, 2020 at 12:38 PM To: Miki Rothschild Yoav Shapira Cc: IG_Well-being_Leads Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") PMM/ Comms felt strongly that it will land poorly to remove this feature that is well appreciated by some when we evaluated the unship in March before COVID surge. I think we might need to re-evaluate if the concerns are even greater now given current comms climate. From: Miki Rothschild Date: Friday, July 24, 2020 at 12:35 PM To: IG_Well-being_Leads Cc: Yoav Shapira Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") FWIW, I think the biggest question is how to land this externally and therefore the primary stakeholders would be our outreach teams. I don't think we have PM bandwidth to quarterback this nor an alternative feature that would make landing this easier. I'm curious to hear your take since you've already dug into this? From: Date: Friday, July 24, 2020 at 11:37 AM To: Yoav Shapira ! Cc: |G_Well-being_Leads Miki Rothschild Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") Hi Team Seems like there's consensus on Time Spent as a good candidate for Unship here. What the Unship central team needs is a commitment to unship, which involves 30 days to scope what it would take to remove the feature and 60 days to execute on the removal. Do we feel like we could move forward with the above timeline? Would we like to investigate more (per Yoav sounds like would be tough given bandwidth?)? Thanks, HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181775 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 4 of 15 Manager, Product Operations From: Date: Wednesday, July 22, 2020 at 9:04 AM To: Yoav Shapira Cc: Miki Rothschild IG_Well-being_Leads Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") +1 to this and also a great opportunity to simplify the app even more (especially for something that is inaccurate). Was also thinking about Vishal's comment about settings being the overflow for features so clearing the way for something more important would be great. If it does stay in there, we really need to differentiate from Archive which is so confusing (both copy and iconography) which I still fumble between. From: Yoav Shapira Date: Wednesday, July 22, 2020 at 8:12 AM To: Cc: Miki Rothschild , IG_Well-being_Leads Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") I remain a fan of unshipping this, too. Previously we thought we might bundle the removal with a new feature around "away mode" or "take a break." But that's been deprioritized, sadly, for another PO. We spent months of eng, DS, and DE time trying to get this accurate in 2018. I defer to their team's bandwidth but I suspect an investigation will be tough. Thanks, on On Jul 22, 2020, at 08:09, Hi Miki/ Yoav/ HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 wrote: META3047MDL-003-00181776 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 5 of 15 Re-surfacing this thread re: unshipping the Time Spent feature. Unship is a program Product Operations runs across IG products (program goal and execution outlined below). Prod Ops leads for this program cc'd on this note. The Unship program just had their IG product group leads meeting. parental leave, but recommended the following: couldn't attend per 1. Understand how inaccurate the feature is (is everything wildly off or could we salvage it by unshipping specific metrics?) 2. If not salvageable, we should unship regardless of comms concerns. We can direct users to the device-level tools since those were not available at time of launch, but he doesn't think PR should block us from moving forward if the feature is not widely used and the feature isn't working. I want to make sure you all are aware this was discussed. The taskforce would like to move forward with unshipping once #1 above is understood. Thoughts? s backfill starts 8/3 and can project manage this investigation, but in the meantime who is the best contact to determine the above? *** Unship Program Goal We protect the user experience and uphold the value of simplicity by evaluating fully-launched features Q/Q and unshipping the ones that didn't have the impact we expected Execution Each quarter, Product Specialists partner with product teams to identify qualifying features, gather metrics, size complexity and impact, and run the final review with the PM Leads for Sharing Experiences, Interests, and Community Miki we haven't yet had a chance to meet, I'm the new Product Operations Manager for Community. I'll set up time for us to meet eachother after this, looking forward to working closer together this half! Thanks, John Manager, Product Operations From: Date: Wednesday, July 15, 2020 at 3:07 PM To: Cc: Subject: FW: feedback requested: unship proposal for Time Spent (in "Your Activity") HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181777 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 6 of 15 From: Date: Friday, March 13, 2020 at 6:23 PM To: Yoav Shapira Cc: Karina Newton Miki Rothschild Charles Sismondo Liza Crenshaw H Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") is the EM POC, in org. From: Yoav Shapira Date: Friday, March 13, 2020 at 6:22 PM To: Cc: Karina Newton Miki Rothschild Charles Sismondo 1 Liza Crenshaw Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") Makes sense. Which Facebook PF team is committed to maintain this after the launch, may I ask? I'd love to ping their eng lead to learn their plan, maybe we can save some technical work. Thanks, Yoav On Mar 13, 2020, at 18:19, wrote: The "product creativity" path seems like a variant of options 3 and 4, where we'd make a decision soon that depends on executing updates TBD. To go down this path, we might ask the new teen well-being team (currently working on social comparison strategy bets) to put some focus here. We could discuss what the tradeoffs are for this path. Re: "Facebook's v2 is unlikely to launch because there's not a team, as you noted. Not in Q2 and not likely in H2 either." HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181778 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 7 of 15 Miki and I have confirmed with Facebook Product Foundation that they are committed to launch v2 in Q2, a decision confirmed today in a review with FB app leads and comm leads. My understanding of why they'd launch this: • They prioritized problematic use in 2019 and mostly finished before dissolving the team. At this point, the feature is done and they see it as a FB comms win. To finish it, they were able to fund product support from those who stayed in other FB Product Foundation teams. It sounds like they'll position it as a part of a general theme of "providing more control to people over how they use FB", which also relates to other work by a variety of FB teams. @Karina here's the working quip for this email if you'd like to suggest alternative framing in the meantime: https://fb.quip.com/fby9AYkPn69i From: Yoav Shapira Date: Friday, March 13, 2020 at 5:38 PM To: Karina Newton Cc: Miki Rothschild Liza Crenshaw Charles Sismondo Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") Yeah, there's definitely space for product creativity. Removing the metrics, pointing out to users how to see them for themselves, while keeping the reminders, etc, are all valid and would address the metric and technical concerns. Thanks, Yoav HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181779 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 8 of 15 On Mar 13, 2020, at 17:11, Karina Newton wrote: Hi, If there's a quip we can pop in, happy to comment/make suggestions on framing but I actually wonder if we can't think about a solution/product pivot that meets the competing goals of outreach and product...or if that conversation has taken place? For policy POV background, the time spent dash + end of feed notification is the biggest proof point we have on tech addiction/problematic use and the tool with the most positive sentiment from our mental health stakeholders -- there's no product work we've done in the last four years that comes close and we wouldn't have the credibility we now have in the social comparison/mental health partner space had we not launched this. It also has outsized policy value as we held multiple design hacks on the time spent tools with the UK, US, and German governments and it was a key subject of Vishal's testimony to UK parliament. In order to land this unship successfully we would need to land the why, and without doing so we would lose significant credibility with our policy and mental health stakeholders (and the fact that people who use it spend less time on Instagram underscores the delicacy). My hunch is adoption is low not only because the device level info is now so good, but also because we did no in-app education about it and haven't evolved or improved it over time (I don't believe we've had PM support since left). The metrics issue is really problematic so I understand why it's so tempting to unship, but I don't think that's going to land well without having something that addresses the underlying issue around problematic use which is a space Well-Being, to point, is so actively working on. If the metrics issue is the core problem, is there not something we could do in terms of pivoting the product where we take away the dash piece that causes metric issues (and can talk about iOS/Android doing such a good job of that) but keep the reminder feature and maybe package/add an entry point to the unfollow/graph trimming pieces growth has done. I.e. is unship our only option or is there some nominal amount of work we could do to evolve it so we can maintain the value we still have in it and explain the why in a thoughtful manner? And obviously, would also like to understand what FB is doing for V2 (which I've been told is happening from the policy side anyway)! From: Date: Friday, March 13, 2020 at 2:21 PM To: Miki Rothschild Yoav Shapira Karina Newton Liza Crenshaw Cc: 1 1 Charles Sismondo HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181780 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") + @Karina, @Liza standing by to help on the proposed re-framing, and adding PR risk (which isn't insignificant). One question that it would be great to answer for the escalation, why is FB focusing on v2 but we aren't? Page 9 of 15 From: Miki Rothschild Date: Friday, March 13, 2020 at 12:58 PM To: Yoav Shapira Cc: Charles Sismondo Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") While I'm generally supportive of unshipping, specifically because I have a sense that the value to users is marginal (given data and OS support) and because of the risks Yoav highlighted (which we should add), I still don't have a good framing for how we'd land this in market. E.g. are we saying that the market perception impact is low or that we just don't care because the former reasons are so strong? I think we need a better way to frame the actual tradeoff between criteria. Additionally, I'll ping to get a sense of their confidence in shipping v2, that's one that we don't need to speculate My suggestion is to frame the criteria in some form of traffic light (e.g. user value, PR risk, effort etc) and add a bit more context on why some portion of the team (i.e. product) believes we should ship while the other doesn't (i.e. outreach) in weighting these factors. From: Date: Friday, March 13, 2020 at 12:19 PM To: Yoav Shapira Cc: Miki Rothschild Charles Sismondo Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") I'm also supportive of unshipping, but one thing for us to think about is how Time Spent could impact future work in the Mental Health area (e.g. Social Comparison) and how we might bring transparency to what people are spending time on (including if they are spending too much HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181781 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 10 of 15 time in the app). Of course there are other ways to do that, but if transparency around time spent in the app is a direction we'd like to explore then revisiting this feature in the future might be necessary. Also to note, while Screen Time is accurate on OS, if turned off (something we don't have insight into), ignored, etc. it might not give us the control of the message we might want to. From: Yoav Shapira Date: Friday, March 13, 2020 at 12:10 PM To: Miki Rothschild Cc: Charles Sismondo Subject: Re: feedback requested: unship proposal for Time Spent (in "Your Activity") Howdy and friends, happy Friday. I'd like to unship this without a replacement. The proposal captures most of why, but here are a couple of missing points. (So in terms of your four options, #1 would be my vote, #3 acceptable but less preferred.) Our data as currently shown is incorrect. It's not just that Apple / Google have better data. Ours is wrong. Far worse. We're sharing bad metrics externally. We've been unable to right it despite several person-months of efforts. Because of that, we've never finished the rollout on Android. This violates our Instagram product principles around holdouts. We can't leave it as-is. So it's wrong (bad enough in itself), can't be fixed easily (we've tried), has been half-rolled-out for a while. Bad. Central External Metrics, the group that audits metrics we provide to the outside world, has called us out on it. Alex Schultz is doing an audit of the Facebook metric, too. (Task with commentary: https://our.internmc.facebook.com/intern/tasks/?t=48697217.) The reason this is relevant is we vouch for these numbers. Any day they're out there is a legal liability. That's why this external metrics group exists, from the video watch time debacle. point of contact there. is a good HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181782 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 11 of 15 We've consistently deprioritized addiction as a work area, which I think is the right call anyhow. Investing further here isn't likely. There's not going to be this option #4 any time soon, I don't think. Facebook's v2 is unlikely to launch because there's not a team, as you noted. Not in Q2 and not likely in H2 either. I think we should decide to unship, and we have a tiny bit of flexibility on timing, but not a ton. Yoav From: Date: Friday, March 13, 2020 at 11:17 To: Miki Rothschild Cc: Yoav Shapira Charles Sismondo Subject: feedback requested: unship proposal for Time Spent (in "Your Activity") We're emailing to start the process of unshipping Time Spent. We think this will be uncontroversial for most product stakeholders given that the phone OSs since replicated the functionality and the metrics in the tool cannot match what is reported at an OS level presenting a metrics integrity risk. Outreach disagrees with unshipping unless we could quickly move on building an acceptable update or replacement. This is one reason to escalate. Adding urgency, the Ads team is working on a history of browser links feature that would fit into the same "Your Activity" entrypoint as time spent, and has offered to unship time spent for us in H1. The decision to unship will affect their design, and are at a critical point this week in either removing the feature or leaving it as is. We're looking for feedback on the framing and recommendation below. We'd also appreciate guidance on how to escalate. If we agree on unshipping in Q2, we have the opportunity to meet with on 3/17 for an expedited unship decision from IG leads via the unship process. Context HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181783 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 12 of 15 In 2018, we launched time management tools on Instagram and Facebook. These were intended to address problematic use and were the flagship launches from IG Well-being at the time. Since that launch, Android and iOS have improved screen time / digital well-being features of their own that provide deeper functionality and visibility into app usage as well as category usage (Social Media for example), as well as the ability to place restrictions on use of apps/categories. There's little to no differentiated value for the Instagram product. Facebook has been working on a v2 of its time spent tools, but dissolved the team in FB Product Foundation in December. It's unclear when in Q2 this product will launch, but they are currently planning a proactive announcement around it. Core Metrics on FB measures usage differently than Instagram, measuring interaction times (for example typing) vs screen time (sitting on news feed without scrolling for 5 min). Core Metrics has asked that we not use their method of counting time, as it is not representative of what 'time spent' that Instagram's feature represents. Unship Criteria Per the unship process and criteria, we evaluate complexity, impact, and reach when considering unship decisions. For the time spent tools: • The feature is of medium complexity as it is its own space, effecting use within the app, but not highly trafficked or visible from major surfaces. Currently, we are not investing in the problematic use space and this feature does not contribute to a team goal. However, because the people who use this feature are actually reducing their time spent (see metrics in appendix), the impact is medium. The reach is low with only 4.6M accounts visiting the tool per day See this document for more details on the evaluation. See additional considerations in appendix Outreach Time Spent tools are crucial proof points when speaking to consumers, the press, and stakeholders about our efforts to combat social media addiction. We've also recently heard from teens that the amount of time they spend on IG is one of the worst aspects of their relationship with the app. Outreach feels strongly that we should not remove these tools (or replace with Ads team proposal, which may fundamentally change the tool's HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181784 Page 13 of 15 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 purpose), and instead invest in updating these tools to better fit user needs, especially since Facebook is likely announcing an update. If that is not an option, potential mitigation includes: 1. Leave tools as they currently exist, as we have not heard any major complaints from external stakeholders or consumers 2. Update the metrics popup messaging to better explain how OS also has these features 3. Explore another way of depicting Your Activity, or add in messaging directing users to view OS tools, but still leave in the Time Management tools. Options 1. Unship in Q2 1. Pros: Ads team will unship the product by end of Q2, app and compute gains made early 2. Cons: Outreach efforts at risk (no current plan for product investment in this space, and Ads proposal may fundamentally change the purpose of the tool to be less about well-being) 2. Leave the feature as is and commit to supporting it indefinitely 1. Pros: Outreach risk mitigated 2. Cons: If we decide to unship later, will need to do the effort ourselves. Well-being isn't investing in problematic use this year. 3. Decide to unship but wait to unship until new teen well- being announcement in H2 as Outreach mitigation 1. Pros: Outreach risk may be partially mitigated (only if new well-being efforts address problematic use) and benefits of unship gained in H2. 2. Cons: Well-being will need to do engineering effort to unship and mitigate. 4. Wait to decide until we see if there's a new teen well- being idea that fits "Your Activity" framing 1. Pros: Outreach risk mitigated by investment in Your Activity strategy 2. Cons: Well-being will need to do engineering effort to unship Recommendation For IGWB Support product and eng, the recommendation is #1: Unship in Q2. Outreach disagrees with unshipping unless we could quickly move on building an acceptable update or replacement. Outreach is HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181785 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 14 of 15 open to unshipping in the future if we have another tangible example of how we are committed to supporting healthy social media usage. Questions: Feedback on framing or recommendation? Considerations we missed? Guidance on how to escalate? *** Appendix: Detail on considerations 1. Well-being strategy: We recently decided to focus our new teen well-being bet on social comparison, giving us conviction that we will not focus on problematic use for the foreseeable future. That said, there are dimensions of social comparison that are amplified by specific types of problematic use, so "Your Activity"-framed features may be considered. 2. New teen well-being bet product ideas: In April, we expect to have early conviction in a strategy for social comparison and a handful of product ideas that are firming up enough to consider execution in Q2. As of early March, we don't have leading ideas that depend on "Your Activity" as a destination. But ideas that connect strongly to "Your Activity" as a destination and concept could surface in future halves. 3. Facebook alignment: Facebook has built a v2 of its time spent tools, with a planned launch in April or May (moved back from March due to coronavirus because they want to optimize for comms impact). IG unshipping at the same time that FB is launching could create a confusing comms narrative. 4. Metrics discrepancy: There's no ground truth measurement for time spent, but there is a 7.2% discrepancy between our internal measurement and OS level measurement on iOS, and a 8.4% discrepancy on Android. The two OS currently does not provide an API to access time spent. 5. Current usage: Everyday, ~4.6M accounts visit the Your Activity page (resulting in about ~6M hits). In an analysis from H2 of 2019, we found that 1.4% DAP have a time spent reminder set, 2.9% people who set a reminder checked the time spent dashboard on a given day, and 25% of people who set a reminder triggered it in a given HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181786 Case 4:22-md-03047-YGR Document 2889-42 Filed 03/27/26 Page 15 of 15 day. If a user triggers their time spent reminder, their next day time spent is reduced by 12%. Overall, the adoption rate for "Time on Instagram" is not high, but once the popup reminder is triggered, it does have an impact on the next day time spent. 6. OS support for similar features: The last 30 days of engagement events show >96% of iOS users are on a version of the operating system that supports this feature, and >54% of Android users. Android has multiple additional apps that will manage "screen time" available in the play store for older android devices. 7. Cost to fix metrics issues: There is no domain knowledge on the feature on the existing team. This would take at minimum a medium lift effort. Even if the metrics are improved, they will never be as accurate as the OS level support, as app crashes, new features, and interrupts such as phone calls can cause time sequences to vanish. iOS and Android have different sources of inaccuracy. Detail on technical limitations here: Time Spent "Your Activity” Status Data Accuracy Update 8. Ads team wants to extend "Your Activity": The Instagram Ads team is hoping to build a history of "Links You've Visited" that replaces the "Time Spent" features within the "Your Activity" tab. They have offered to unship the product in H1 as part of their effort. 9. App Performance and Compute cost: After discussion with Core Metrics, app cold/warm start and server compute will improve if we unship the feature. The amount of improvement is unclear. 10. Outreach: see above HIGHLY CONFIDENTIAL (COMPETITOR) META3047MDL_RVOL003 META3047MDL-003-00181787