Internal Facebook docs show time spent by US teenagers was down 16% YoY, young US adults were also spending 5% less time on the social network, and more
Millennials are losing interest in Facebook too, new documents show https://www.bloomberg.com/... via @technology Kurt Wagner / @kurtwagner8 : NEW story w/ @damclaugh: Internal docs and research show Facebook isn't just losing popularity with teens. It's losing popularity with those under 30. They're spending less time, creating fewer posts + sending fewer messages. https://www.bloomberg.com/... Carl Quintanilla / @carlquintanilla : The internal documents showed “a troubling trend that seemed to be accelerating: Facebook was losing popularity with teens and young adults.” (via @business) $FB https://www.bloomberg.com/... https://twitter.com/... Kurt Wagner / @kurtwagner8 : FB's teen issues have been well documented. But historically, most teens who joined IG would “age up” to Facebook by age 19 or 20. That's no longer happening. Internal research shows “aging up” is taking longer (e.g. 24 or 25 years) old if at all From an internal report: https://twitter.com/... @jedimarkus77 : Forgive me for sounding not-bearish on the Book of Face but hasn't this been a narrative for a looooooooong time? Maybe the acceleration rate is the new story... Zuck can just claim “transitory” right? $FB https://twitter.com/... Michele Boldrin / @micheleboldrin : Not a good week for FB. I hope it gets worse and worse week after week :) https://www.bloomberg.com/... Martin Sfp Bryant / @martinsfp : “The declines among teens and young adults on Facebook have been almost invisible to outsiders. Facebook doesn't break down its user numbers by age group, and it has evaded questions from Wall Street analysts” https://www.bloomberg.com/... Lisa Abramowicz / @lisaabramowicz1 : Researchers found that “time spent” by U.S. teenagers on Facebook was down 16% year-over-year, and that young adults in the U.S. were also spending 5% less time on the social network. https://www.bloomberg.com/...
Context & Ripple Effects
This is the internal-document confirmation of an erosion that outside measurement already flagged twice: a 2014 study showing the teen share of social media users slipping on Facebook and a 2017-to-2019 Edison Research estimate of 15M fewer US users, concentrated in ages 12–34. What the new documents add is velocity — a 16% year-over-year drop in US teen time spent, 5% among young US adults, and evidence that the historical 'aging up' pipeline into Facebook is starting late or not at all.
The same-day reporting shows the company internally describing this as 'accelerating,' which matters because Facebook's revenue model was built on the assumption that each cohort would eventually arrive. A cohort that never arrives is a permanent hole in the audience base, not a delayed one.
First-order effects
- US advertisers buying against under-30 audiences face shrinking engagement inventory on Facebook right now — fewer hours, fewer posts, fewer messages per user — while the company can no longer present youth usage as flat or growing.
Second-order effects
- The competitive pressure lands on Meta as a forced product response: the coverage shows it later revamping the feed to surface posts beyond friends and family specifically to fight TikTok for younger attention, trading its core social-graph premise for reach.
- Rival platforms compete for the same finite US teen attention pool, so every hour lost by Facebook is a bidding-war hour gained elsewhere — pushing up the cost of reaching young users across all networks.
Third-order effects
- If the aging-up break holds across cohorts, Facebook structurally becomes an older-skewing network in its largest ad market, splitting the industry between youth-attention apps and an aging incumbent — with advertiser budgets following the age split rather than total scale.
The trend: US social media is bifurcating by age, with Facebook's own research confirming that the youth attention it once absorbed by default now has to be won back feature-by-feature against newer rivals.