The Attention Fracture

For roughly seventy years, American culture had a schedule. Network evening news ran at 6:30 p.m. The Sunday-morning shows ran on Sunday morning. Late night ran late at night. Films opened on Fridays. Magazines arrived on Tuesdays. The schedule was not just a media convention — it was the temporal scaffolding of public life. A meaningful share of the country watched the same things at roughly the same time, encountered the same advertisements between segments, and woke up the next day with overlapping mental furniture. That scaffolding has now collapsed. Streaming surpassed combined broadcast and cable viewing in May 2025 (Nielsen, "The Gauge"). Short-form video has displaced television as the dominant medium for adults under 30. News reaches younger audiences primarily through creators and platforms rather than newsrooms and brands. The shared schedule is gone, and the attention economy that replaced it operates on completely different physics.

The result is what this report calls the attention fracture: a condition in which two adults living in the same household, working at the same company, voting in the same district, and earning similar incomes can spend their media hours in almost completely non-overlapping content universes — and in which the marginal hour of attention is increasingly contested by algorithmic feeds, creator ecosystems, and on-demand libraries rather than by scheduled, editorially curated programming. The consequences cascade through news, brand strategy, political communication, civic ritual, and the basic experience of having a culture in common.

The End of the Schedule

Nielsen's "Gauge" report has tracked how Americans allocate their television time across broadcast, cable, and streaming since 2021. The crossover point arrived in late 2023, when streaming first passed cable individually; by June 2025, streaming captured 46% of all TV viewing time, against 22% for cable and 18% for broadcast. The remaining 14% is split among gaming, physical media, and unmeasured uses. These percentages tell a clear story about delivery, but they understate what has changed about consumption.

Inside the streaming bucket, viewing has fragmented further. YouTube alone now accounts for roughly 12% of all U.S. television time — more than Netflix (8%), Disney+ (2%), and Hulu (3%) combined when measured on connected TVs. Roughly 45% of YouTube's viewing time on TV screens is creator-uploaded video, not professional studio content. The implication is that even on the device historically synonymous with mass-produced, professionally edited entertainment, a substantial and rising share of attention is being captured by videos made by individuals without studios, distributors, or editorial supervision.

Off the television, the shift is more pronounced. eMarketer's 2024 time-spent estimates put U.S. adults at 4 hours 25 minutes per day on digital video (up from 2 hours 15 minutes in 2019), with mobile capturing two-thirds of that time. TikTok alone averages 95 minutes per active user per day among U.S. adults under 35; YouTube Shorts, Instagram Reels, and Snapchat Spotlight collectively capture a comparable amount among the same cohort. For users under 25, the median day now contains more time spent on short-form vertical video than on every form of long-form professional video combined.

News in the Feed

The news consequences of the attention fracture are the most thoroughly documented and the most politically consequential. Pew's 2024 News Platform Fact Sheet finds that 54% of U.S. adults at least sometimes get news from social media — up from 27% in 2013. TikTok is the fastest-growing source: 39% of adults under 30 now regularly get news on the platform, up from 9% in 2020. Facebook and YouTube remain larger in absolute terms but flat or declining among the under-30 cohort. The migration is generational and directional.

Pew's November 2024 study on "news influencers" added a critical layer: 21% of U.S. adults regularly get news from individual creators on social platforms, rising to 37% among adults under 30. Among that under-30 cohort, the share who say creators help them understand current events (74%) is statistically tied with the share who say the same about professional news organizations (72%). For the median young adult, a podcast host, a TikTok explainer, or a Substack writer is now a cognitively equivalent news source to a national newspaper.

Reuters Institute's 2024 Digital News Report adds the international confirmation: across 47 markets, 56% of news consumers under 25 say they prefer to access news through "side-door" routes (social media, search, aggregators, push notifications) rather than by visiting a news brand directly. Direct visits to news websites among this cohort fell from 32% in 2018 to 17% in 2024. The branded news product — front page, masthead, edition — is becoming a back-end production facility for content that audiences encounter elsewhere, unbundled, and increasingly without the brand attached.

The downstream effect on news economics is severe. As distribution moves from owned properties to feeds, the ad revenue and subscription revenue that funded professional newsrooms migrates to the platforms doing the distribution and to the creators producing platform-native content. The Reuters Institute estimates that U.S. newspaper newsroom employment fell 57% between 2008 and 2023, with the steepest declines concentrated at metro and regional outlets. Local news deserts now cover roughly half of all U.S. counties.

The Algorithmic Substitute for the Editor

The functional replacement for the editor — the human who decided what the audience would encounter and in what order — is the ranking algorithm. TikTok's "For You" page, Instagram's recommendations, YouTube's home feed, X's "For You" timeline, and Spotify's autoplay are all variants of the same architecture: a machine-learned predictor of what each user is most likely to engage with, optimized continuously against engagement signals, with no editorial intent and no commitment to balance, completeness, or shared experience.

The first-order effect of this substitution is the decoupling of cultural relevance from professional judgment. A song, video, or creator can reach tens of millions of people without ever passing through any institutional gatekeeper. The second-order effect is the disappearance of the shared cultural experience itself. Two people scrolling TikTok in the same room are looking at almost entirely different content; the platform's value proposition is precisely that it knows each user well enough to construct a personal feed. The shared schedule has been replaced not just with on-demand consumption but with content streams that no two people share.

The political and civic consequences of this architecture are still contested. Empirical work by Bail, Allcott, and others has consistently found that algorithmic feeds increase exposure to extreme content but do not, by themselves, account for most of the observed polarization in American politics — the more powerful driver appears to be the broader incentive structure of attention markets, in which emotionally charged, high-conflict content outperforms calmer alternatives regardless of algorithmic curation. The algorithm is an amplifier of an underlying economic logic, not the source of it.

The Creator Economy as Infrastructure

Goldman Sachs estimates the global creator economy at $250 billion in 2024, projecting growth to $480 billion by 2027. SignalFire estimates that 50 million people worldwide now identify as creators, with roughly 4% earning enough to consider it a primary income source. In the United States, the top tier of full-time creators (roughly 200,000 individuals) now earn aggregate revenue comparable to the entire newspaper industry's editorial budget.

These figures matter less for their absolute scale than for what they imply about distribution of cultural authority. The creator economy is not a cottage industry adjacent to the media business — it is the media business, increasingly, for audiences under 35. Brand strategy, political communication, public-health messaging, education, and entertainment are all being rebuilt around named individuals with audience relationships rather than around institutions with reach.

Three structural features of the creator economy are underappreciated. First, creators function as quasi-publishers with no editorial overhead — they bear distribution risk on platforms they do not control, and their relationship with audiences is platform-mediated and revocable. Second, audience trust in named individuals consistently outperforms audience trust in the institutions that house equivalent content (see "The Trust Recession" report). Third, the creator economy operates on attention as both raw material and finished product — the work is the audience relationship, and the content is the means of maintaining it.

The Return of Audio

Within the broader fracture, podcasting has emerged as a countervailing pattern: long-form, often slow-paced, deliberately chosen rather than algorithmically served. Edison Research's 2024 Infinite Dial finds 47% of Americans 12+ listen to podcasts monthly (158 million people), up from 24% in 2017. Average weekly time spent listening among podcast consumers is 8 hours and 15 minutes — comparable to broadcast television viewing among the same age cohort.

Podcasting matters analytically because it represents the only major medium gaining audience without depending primarily on algorithmic discovery. Listeners subscribe deliberately to a small number of shows and listen to most of each episode. The economics favor depth over breadth and named hosts over institutional brands. Joe Rogan, Lex Fridman, Alex Cooper, the Daily, Pod Save America, This American Life, Serial — the most successful podcasts are either named hosts or cohesive editorial brands led by identifiable humans, not network-style programming blocks.

The migration of substantive interview, debate, and explanatory content from broadcast and cable into podcasting and long-form YouTube has reshaped the public-square function. A two-hour unedited conversation now reaches audiences comparable to a prime-time cable news segment, with a fraction of the production cost and almost no institutional oversight. The most influential long-form interviews in the United States in 2024 — across politics, business, and culture — were largely conducted by independent hosts on YouTube or podcast feeds, not by network anchors.

Streaming Subscriptions and the Bundle Returns

On the on-demand side, the streaming wars produced an outcome that few predicted at the start: audiences have re-bundled. Antenna's 2024 State of Subscriptions data shows that the average U.S. streaming household pays for 4.3 services, with churn rates above 40% annually for non-Netflix providers. The consumer behavior is clear — subscribe, binge, cancel, rotate. The economic response from providers has been to recreate the cable bundle in new form: ad-supported tiers, multi-service partnerships (Disney+/Hulu/Max, Verizon's combined offerings), and platform aggregators (Amazon Prime Video Channels, Apple TV's centralizing role).

The implication is that the post-cable world is not a permanent state of fragmentation but a transitional period toward a new bundling equilibrium. The unit of bundling, however, is no longer the channel — it is the service, the ad-tier, the platform, and increasingly the creator. Spotify's all-audio strategy, YouTube Premium's video-plus-music-plus-podcast bundle, and Substack's creator-bundle experiments are early indications of where the recombination is headed.

The Attention Cost

The most-cited claim about contemporary attention — that average attention spans have collapsed to 8 seconds — is methodologically weak and frequently misattributed. The more rigorous evidence comes from the work of Gloria Mark and colleagues, whose lab and field studies on workplace attention show that the average duration of focused attention on a single screen task fell from roughly 2.5 minutes in 2004 to 47 seconds in 2023. The mechanism is not biological but environmental: the proliferation of interrupting devices, notifications, and rapid-cycle feeds has reshaped the practical conditions under which sustained attention is possible.

The downstream effects on civic and creative life are still being measured. Reading time per capita has fallen across every age cohort under 60 in the American Time Use Survey since 2010. Long-form journalism, scholarly writing, and book reading occupy a smaller share of the median educated American's day than they did a decade ago. Whether this represents a permanent shift in the cognitive economy or a transitional period in which sustained attention will re-emerge as a luxury skill is an open empirical question. Several plausible signals — the growth of long-form podcasts, the resurgence of book sales among Gen Z, the success of slow-news products like Tortoise and Semafor's longer formats — suggest that demand for depth has not disappeared. It has re-bundled into new formats and is competing on different terms.

What the Fracture Means for Brands

For commercial brands, the attention fracture has rewritten the rules of media planning. Reach campaigns that historically depended on a small number of mass-audience properties (network prime time, Super Bowl, large magazines, top-100 radio) now require dozens of platforms, formats, and creator partnerships to approach equivalent unduplicated reach. The cost per thousand for true mass reach has risen sharply, and the operational complexity of running a unified campaign across the modern media stack has risen even faster.

Three patterns are emerging. First, brands that succeed in the fractured environment increasingly run a barbell strategy: a small number of premium, long-tail brand investments (sponsorships, original content, live events) paired with a broad creator and performance layer that operates at high frequency and is measured on direct response. The middle layer — generic ad placements across mid-tier media — is collapsing economically. Second, the most effective brand work in the new environment is increasingly unbranded in form: it travels as content, not as advertising, and derives its credibility from the named creators who carry it. Third, brand-built audiences (newsletters, owned communities, proprietary apps) are becoming defensive infrastructure against the platform tax, with the best-managed brands now treating their first-party audience the way newsrooms treat their subscriber base.

What Comes After the Fracture

The attention fracture is not a stable end state. Two opposing forces are now visible in the data. On one side, continued platform consolidation, AI-generated content scaling, and the feed-ization of every remaining medium (including news, books, and education) push toward further fragmentation, deeper personalization, and a continued decline of shared cultural experience. On the other side, fatigue with algorithmic feeds, rising willingness to pay for curated and human-edited products, and the slow recovery of long-form audio and writing point toward a partial re-bundling around trusted humans, named editors, and deliberate consumption.

The most likely medium-term outcome is neither a return to the shared schedule nor a permanent state of total fragmentation, but a stratified attention economy in which a thin layer of deliberately chosen, deeply trusted, named-human media products sits on top of a vast, algorithmic, platform-mediated layer of ambient content. The former will increasingly carry the substantive weight — the news that informs civic action, the analysis that informs decisions, the cultural products that shape identity — while the latter will provide the constant ambient consumption that fills the rest of the day. The question for brands, institutions, and creators alike is which layer they are operating in, and whether their economics support that position.

The shared schedule that organized twentieth-century cultural life is not coming back. What is being built in its place is a cultural environment in which attention is the scarce resource, named humans are the trusted intermediaries, and the unit of cultural production is increasingly smaller, faster, and more personal than at any point in the broadcast era. The institutions that thrive in this environment will be the ones that understand attention not as a measurable output of media spend but as the underlying resource that all other forms of cultural and commercial influence depend upon.