How should we default new teens into new interactions settings? (presentation)

Raw text

Transcribed text as recorded, including headers and formatting fragments.

Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 1 of 25 AMENDED Exhibit 336 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 2887-13 Filed 03/27/26 DOCUMENT SLIPSHEET Bates Number: META3047MDL-019-00092508 Custodian: CASTANEDA DIEGO Author: Filename: HOW SHOULD WE DEFAULT NEW TEENS INTO NEW INT 16RB2D6KTXZSYPGTC7FOMHWFPWXXDG9DDSE9CF8PGS6E.PPTX Create Date: 5-28-2020 12:00 AM Last Modified Date: 6-2-2020 12:00 AM TNAG BEGBATES: METATNAG-005-00092508 TNAG ENDBATES: METATNAG-005-00092508 MAAG BEGBATES: METAMAAG-008-00092508 MAAG ENDBATES: METAMAAG-008-00092508 depobook.com EXHIBIT META- Castaneda-16 Page 2 of 25 Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 3 of 25 How should we default new teens into new interactions settings?: a survey of safety and value (Data Science) Diego Castaneda (UX Research) Should we default new teens into a safety settings? Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Restricting Reachability 341 Account Privacy Choose who can see what you share του tan change this anytinesting T-Mobile < 8:05 PM Privacy 50% Interactions Comments Private Orly as yo you post and imara you Public Anyone can yaω τις επανάται Tags @Mentions +) Story Activity Status Messages Connections Account Privacy Restricted Accounts Blocked Accounts Muted Accounts -Close Friends Page 4 of 25 Ultimately privacy is more about content visibility so we decided to think of reachability... we should not mix the two.. Privacy is unrelated to reachability.. Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 5 of 25 TL;DR Premise: "Building a safer account model is the right thing to do. Defaulting people into new, restricted settings could protect people from unwanted interactions, especially from people they don't know." >> After an initial analysis challenged this premise, we decided to invest additional DS and Research resources into assessing old and new signals. Principles. We needed to closely examine the relationship between providing user safety (protecting people from unwanted interactions) and delivering value to users (healthy social engagement). How could restricted interaction settings increase safety Restricted interaction settings can shield people from Ul. DMs are where most unwanted interactions (UI) happen, especially for new teens users. Starting new teens in a default restricted reachability setting significantly reduce the volume of overall unwanted interactions in DMs. How could restricted interactions settings sacrifice value Data projection show a strong potential for loss of valuable interactions in DM's. People want meaningful interactions from both people they know and don't know. Fostering positive relationships may mitigate the development of bullying behaviors for teens, in particular. People value consistency and predictability in privacy related settings. People are unlikely to change settings without interventions. Constraints.. Consider talking about these later. Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 6 of 25 Recommendations Don't change the current default interactions model: doing this would cause people to miss out on too many valuable interactions. Better understand how restricted reachability settings for DMs, tags, and mentions provide value for people who opt in. We need to understand who these people are and learn how we can encourage their adoption of these settings. Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 7 of 25 Motivation Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 8 of 25 Context The IG Bullying team set a goal: 'build a safer account model in H1 2020' . We knew that people experience unwanted interactions involving random strangers and people they are relatively unconnected with. People experience unwanted interactions in tags, mentions, and DM messages &/or requests. Since we started working on this, the IG Bullying and Direct teams have tested new settings giving people more control over who can interact with them in tags, mentions, and DMs. Key Questions . Should we create a new account model by defaulting people into restricted interaction settings for tags, mentions, and/or DMs? • If so, for which people? • What should be the default setting(s) How do we use data signals to understand 'unwanted interactions' in a way that accurately reflects people's experiences? How do we maintain meaningful interactions while preventing unwanted interactions? Speak to this more Goals and th question We went from "knowing" a problem existed.. maybe .. and then going towards really concrete settings based on an abstract problem.. How the f do we get from hte problem that we know is real and one that is based on a product solution... we were making decisions about defaulting people into settings that we had not tested yet.. Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 9 of 25 Principles: Assessing Safety and User Value Value and Safety are both important. Safety: protecting people from unwanted interactions Value: facilitating healthy social engagement Q. How do Safety and Value signals inform our POV on defaulting users into restricted settings? Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 DS and UX Collaboration . DS and Research worked together. Here's what we did: Developed an analytical framework. When starting with an abstract premise, it's important to develop an analytical framework to aggregate insights to inform decision-making. Synthesized analysis and research signals, qualitative and quantitative. Followed the data. Challenged assumptions. Final recommendations contradicted initial product intuition. Page 10 of 25 What went well.. What could be improved.. We are answering a larger conversation...multiple places.. Blind spots we can each help other with. Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 11 of 25 How could restricted interactions settings increase safety? Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 12 of 25 Restricted interaction settings can shield people from unwanted interactions. Restricted interactions settings are highly valuable for a small minority of users who try them. What % of users' tags & mentions are prevented by non-default settings? Mentions Tags # of users 0% A 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Percent of all interactions prevented by non-default setting The number of users who who Source: public test of tags and mentions settings (NZ, AU) Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 13 of 25 DMs are where most unwanted interactions happen . . 7% of teens report experiencing bullying on Instagram each week. (H1 2019) According to NETS user survey, 40% of all bullying experiences are reported in Direct Messaging. B&H prevalence in DM's is very high (10%) according to the September prototype metric Severe violations targeted to those under 18 are known as a high prevalence problem in DMS PERCEIVED REACH: SURFACE BREAKDOWN Explore and Profile are each the most common surface for 5 out of 16 negative experiences. In addition, Direct is the most common surface to encounter suspicious accounts, and to experience bullying and solicitation. People most commonly see drug sales and SSI in Stories, and see spam in Feed. eating disorder terrorism misinformation child safety violence hate nci firearms nudity drugs bullying seeing sai spam commercial solicitation Where did you see... kiny Food bullying experiencing suspicious account "We didn't ask the surface follow-up for eperenong a compromised account or authentic angagement Source: NETS Survey October 2019 Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 14 of 25 回 What if, by default, new teens couldn't receive DMs from people they don't follow? Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 15 of 25 ~53% Teens delete a lot of DMs from people they don't follow in their first four weeks on Instagram. of deleted DM threads & requests are from people the recipient doesn't follow. Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 16 of 25 Deletions are a strong indication of unwanted interactions 1~13% of new DM threads & requests deleted by the new users result in the recipient/new user blocking the sender (vs. ~6% of all DM threads & requests) 2 ~99% of all blocks involving a pair of users (new teen and sender) who had some DM interaction in the new user's first few weeks are associated with a deletion. Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 17 of 25 Teens also KEEP a lot of DMs from people they don't follow in their first four weeks on Instagram. ~53% ~55% of deleted DM threads & requests are from people the recipient doesn't follow. of kept DM threads & requests are from people the recipient doesn't follow. So, defaulting new teens into a restricted setting could eliminate up to half of both wanted and wanted DM messages. This is a difficult tradeoff to navigate. Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 18 of 25 How could restricted interactions settings sacrifice value? Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 19 of 25 Data projections show a strong potential for loss of valuable DM interactions if we make a drastic change for existing teens. 10 million 9 million Estimated impact of smart defaults on DMs: worst-case & best-case scenarios for TEEN accounts 9.5 million 100% wanted, 100% keep, # unwanted 100% wanted, 100% keep, # wanted 10% wanted, 100% keep, # unwanted 10% wanted, 100% keep, # wanted DM threads preempted per day 8 million 7 million 6 million 5.4 million 5.4 million 5 million 4 million 3 million 2 million 1 million Worst Case Best Case 950 k % Non-deleted DMs "wanted", no negative feedback It depends on what % of kept DMs (from people you don't follow) are actually 'wanted'... For existing teen users, earlier analysis produced the following two scenarios of DM impact, assuming that we were to default all of these accounts to be reachable in DMs by only people they follow and that 100% of teens keep the change. The 'worst case' corresponds to the assumption that 100% of non-deleted DMs are 'wanted', while the 'best case' assumes that only 10% of these are actually wanted. If we assume that If we flipped the switch overnight on all teens into a resticted setting, the potential for loss of daily DM's Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 20 of 25 Teens want meaningful interactions from both people they know and don't know and this contributes to their social graph "A girl that I knew from school messaged me on Instagram. I ignored it at first because I didn't know who it was but then I figured out who it was, If I limited who could DM me then I would've never seen this." "I get a lot of spam DMs, but I can't just ignore them... Just last week someone from my middle school found me on Instagram and DMed me. I wouldn't want to miss things like this." Connecting with new friends via DM. Data shows that approximately 50% of 1:1 DM requests sent to new users in their first fourteen days on Instagram lead to the sender following the recipient at the end of the first 4 weeks. Reconnecting with an old friend via DM's. "I don't want random accounts to tag me, but at the same time it could be a friend or someone at school that I just haven't connect with on Instagram yet" Connecting via tags. Source. Bullying concept testing with teens (Hitomi) Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 21 of 25 Fostering positive relationships can mitigate the development of bullying behavior in teens. Experts on teenage brain development describe the four qualities of the teenage brain that propel them through adolescence. Novelty seeking Increased emotional intensity Creative exploration Social engagement "By using a lens that takes into account the complex social nature of adolescents as well as their unique neurodevelopmental stage, bullying prevention can be viewed in a new way...Bullying prevention efforts should meet young people where they are." Social Psychologist and School Counselor Dr. Julie McDaniel Source. A New Lens for Bullying (Diego) an expert on teenage brain development, who describes the four qualities of the teenage brain that propel them through adolescence. • Novelty seeking • Increased emotional intensity • Creative exploration • Social engagement Instagram is uniquely positioned to "inspire young people at a psychophysiological level to strengthen their peer connections and to develop new peer relationships. Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 22 of 25 People value consistency and predictability in privacy related settings. • Consistency, Predictability and Control are super important for all users. • We need to meet people's expectations with respect to private or public setting. In research studies, some suggested that interactions controls be set to "people they know" for private accounts and "everyone" for public accounts. "What if I'm reassessing the purpose for my account? This would include rethinking what all interaction look like. For example, I've decided I want to go public, maybe this means I want to have more interactions, so opening [tags and mentions] to "everyone" makes sense and is how I can get the most out of this." - Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 23 of 25 People are unlikely to change their interaction settings on their own. • In the recent Better Interactions test, only 4% of views of tags and mentions entry points result the user clicking through to view the new settings options. Of these, about 15-20% result in a settings change ultimately, only about 0.6-0.8% of users encountering these entry points actually make a change. Ultimately, ~0.5% of DAP in the six countries involved in the public test of tags and mentions settings adopted the change during the test period. • Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 24 of 25 Recommendations Case 4:22-md-03047-YGR Document 2887-13 Filed 03/27/26 Page 25 of 25 Recommendations • Don't change the current default interactions model: doing this would cause people to miss out on too many valuable interactions. 0 The value of DM's in promoting wanted social interactions forms key pieces of a healthy Instagram experience and social graph growth in teens. 0 Future product decisions should respect the value of DM interactions between unconnected users. Restricted reachability in DMs, tags, and mentions can provide value for some people who opt in. • Although, unwanted tags and mentions are a less serious problem for an overwhelming majority of users, we should still be proactive in meeting user needs and mental models of safety. While we feel good about balancing towards value, we can take proactive steps to honor when users desire safety. Going forward Conduct user research to understand public private user journey in first 14 days. Upsell intermediate better interaction controls to segments of users who would benefit most (analysis TBD) Explore heuristics of DM's to determine better definitions of unwanted vs meaningful interactions. Explore product solutions to add friction to unwanted, while still allowing for interaction.