Late Night Use: IG Mental Well-being, Problematic Use • Strategy and Design

Raw text

Transcribed text as recorded, including headers and formatting fragments.

Case 4:22-md-03047-YGR Document 3029-7 Filed 05/08/26 Page 1 of 10 Exhibit 104 Case 4:22-md-03047-YGR Document 3029-7 Filed 05/08/26 Page 2 of 10 Late Night Use IG Mental Well-being, Problematic Use • Strategy and Design • [A/C Priv] 21 Jul 2022 Case 4:22-md-03047-YGR Document 3029-7 Sub-area Filed 05/08/26 Page 3 of 10 Problematic Use Teen prevalence Problem severity Instagram coverage Tik Tok coverage Why Late Night? In collaboration_with CWB, we identified 4 critical PU sub -areas related to potentially problematic on -platform behaviors Of these 4, we identified late night use as a top priority for H2 2022 as negative impact of poor sleep is well -documented (ref) and this is where we have coverage gaps for ~8% of teens Late night use Late night usage/during times that are likely displacing sleep for most people (12am- 4am) 7.4% of teens WAU (over one week) Medium Not supported exhibit one or more late-night problematic behaviors Frequent use Habitual and repeated checking of: app, notifications, direct, quick switching between surfaces Sizing in progress Long periods of use Spend more time in the app than intended, lose track of time, low awareness of time spent and long sessions Passive use Sessions with low/no engagement, long scrolling 33% of teens WAU Have at least one 20min+ session per week 9% of teens WAU Scrolls with 35+ impressions on the same topic Medium Not supported High Medium Time-spent awareness videos Auto-silenced notifs after 9pm Night insights App-opens insights Daily usage limits Take a break (upsell and tool) Take a break (tool only, no upsell) Alternative topic nudge Usage- awareness videos Case 4:22-md-03047-YGR Document 3029-7 Filed 05/08/26 Page 4 of 10 Teens (13-17) late-night use Teens late night usage (12am - 4am) 43.3% (77.9M) WAU have at least 1 late night session over one week 4.6% of WAU teens have 1 or more late sessions every night over one week 7.4% (13M) WAU exhibit at least 1 late night behavior associated with problematic use (see proxies below) over one week Problematic behavioral proxies (p95) WAU 18 min+ spent on IG between 12 am and 4 am 5 sessions between 12am and 4am 1 session between 12am and 4am with 18+ surface switches Late Night PU metrics identified as See FCI Countries breakdown here. behavioral proxies (thresholds derived from p95). Case 4:22-md-03047-YGR Document 3029-7 Filed 05/08/26 Page 5 of 10 • Previous research which linked problematic Facebook use to late -night use, has been associated with both later bedtimes and rising times and with insomnia. Because of the association between PU and lack of sleep, the fraction of sessions that occur late at night (12AM was used for this paper. 4AM) Why is 12AM - 4AM late 12AM-4AM is the approximate time period associated with higher negative affect in Golder & Macy 2011 night? • On TikTok all teens do not receive notifications at night by default: 13-15 yrs old - notif muted starting at 9PM 6-17 yrs old - notif muted starting at 10PM Apple defaults Focus Mode/Do Not Disturb to 10PM 7AM daily Case 4:22-md-03047-YGR Document 3029-7 Filed 05/08/26 Page 6 of 10 We are currently collaborating with CWB on a “ground truth” measurement survey of problematic use on Instagram, and expect to have a PU prevalence estimate by end of H2 (see understand plan). In the meantime, we can leverage behavioral proxies identified in a Facebook study to estimate PU on Instagram. That study found that 12.5% of FB app users worldwide experience PU based on user self- report (ref) and identified the following metrics as PU proxies: PU Behavioral Proxies 1. No-engagement late night sessions 2. Late night high volume product switching 3. No-engagement repeat sessions (Less than 10 min from end of prior session) 4. High volume of short sessions (<15 sec) 5. High Frequency notifications checking Using these signals, plus late night long sessions (since displacing sleep is one ofthe negative life impacts ofPU), we estimate that 18.3% of IG teen users exhibit behaviors associated with PU. Case 4:22-md-03047-YGR Document 3029-7 Filed 05/08/26 Page 7 of 10 Late Night Problematic Behavioral Proxies 1. Late night sessions Avg total Stated teen/non- ! teen sessions late night P50 total sessions late night P75 total sessions late night P80 total sessions late sessions late! T 1 P95 total night night non-teen teen 0.9 0.0 0.7 1.0 1.1 0.0 0.8 1.1 4.0! 5.3 1. Late night high volume product switching Late night use x teens (13-17) L non-teen teen Stated teen/non- teen Avg product switches late P50 product switches late night night P75 product switches late night P80 product switches late P95 product switches late night night L 3.8 0.0 3.1 4.4 17.6 1 3.7 0.0! 2.8! 4.0 18.5! 1. Late night long sessions (since displacing sleep is one of the negative life impacts of PU) Avg Late night P50 Late night P80 Late night P95 Late Stated teen/non- ! P75 Late night teen time spent (min) time spent (min) time spent (min) time spent (min) night time spent (min) non-teen 3 0 3 5 17 teen 3 0! 3 4! 18! Case 4:22-md-03Late Night Problematic Behavioral Proxies by FCI Countries: With p95 set by the global average this is the % of WAU by Country(FCI): % of total WAU teens with at FCI Country US Australia Brazil France Germany India Japan Korea UK least 1 PU behavior #WAU Teens with at least 1 PU behavior 11.74% 1,522,833 6.27% 73,118 8.10% 861,713 12.32% 256,601 5.14% 107,208 3.05% 1,436,068 7.01% 427,045 13.49% 280,498 4.82% 106,842 Late night use x teens (13-17) This could indicate that there are key differences in markets in terms of either norms or problematic use, which remains to be seen by further research. If we were to set thresholds based on the market it would like this: Late Night Total FCI Country: Sessions (Teens) p95 (Teens) p95 Late Night Minutes spent Late Night Product Switches (Teens) p95 % of Total WAU teens with at least 1 PU behavior Global 5 18 18 7.4% US 10 19 13 7.1% Australia 5 12 13 7.8% Brazil 5 22 24 7.6% France 9 19 13 8.0% Germany 5 10 12 7.4% India 2 11 34 6.6% Japan 6 10 13 7.7% Korea 10 20 15 7.3% UK 5 9 12 7.5% Case 4:22-md-03047-YGR Document 3029-7 Filed 05/08/26 Page 9 of 10 Sizing/Severity Late Night Use • 43.3% (77.9M) of WAU teens have at least 1 late night session* L7 • WAU teens L7 with at least 1 session: Ο Ο >10 min: 13.6% (24.5 M) >30 min: 3.24% (5.8M) >1 hr: 0.59% (1M) • 67M Sessions (0.4% of total teen sessions) >10 min late night L7 for teens 4.6% of WAU teens have 1 or more sessions every night L7 • 33% of teen sessions between 12 am and 4 am start due to a notif (26% receive and click after midnight) *Late night timeframe defined as 12 am to 4 am, sessions > 0 seconds Case 4:22-md-03047-YGR D Filed 05/08/26 Page 10 of 10 Document 3029-7 Filed 05/08/26 For Users with Late Night PU Behaviors: • • An average of 40% of teen's sessions, start from a notification or an app badge. 18% of dau start a session from a notification that was sent and clicked within 12 am and 4 am Π 43% are direct text notifications Users who start a session organically have a higher daily average time spent in reels, compared to users who start a session from notifications/app badge whose daily average time spent is higher on direct. Sizing/Severity Late Night Use Session start attribution is organic Daily avg Daily mean Daily p95 Session start attribution is notification or app badge Daily avg Daily mean Daily p95 Surface minutes: minutes : minutes Surface:minutes:minutes minutes reels 8.0 2.0: 36.0 home 6.4 3.1 23.8 home 3.9: 1.4: 16.5: direct 9.9: 2.3: 47.3: reels 6.8: 1.9: 29.9: profile 3.5: 0.8: 15.9 explore 5.7: 0.7: 29.1 profile 3.7 1.3: 15.6: saved 3.3 0.7 15.1 Stories creation 2.7 0.9: 11.9: direct 2.4: 0.4: 10.7 saved 2.5 0.6: 11.6: