The Attention Landscape · established evidence
Two Crossovers, Twelve Years Apart: Digital-over-TV (2013) and Streaming-over-Linear (2025)
Two dates, twelve years apart, mark the clearest regime changes in modern media consumption. In 2013, US adults' daily time with digital media first exceeded their daily time with television, a shift eMarketer projected at 5 hours 46 minutes digital versus 4 hours 28 minutes TV for the following year, with mobile as the specific driver. In May 2025, Nielsen's The Gauge recorded streaming exceeding broadcast and cable combined for the first time, 44.8 percent to 44.2 percent of US TV usage, then reaching a new high of 47.5 percent by December 2025. Both crossovers were measured by industry-standard panels, both were dated events rather than gradual drifts, and both rewarded advertisers who reacted early over those who waited for the trend to become obvious.
The first crossover: digital over TV, 2013
The 2013 crossover is one of the most widely re-cited reference points in media planning, and for good reason: it marks the moment an entire category, all of digital media combined, overtook another entire category, television, in daily adult time spent. eMarketer's reporting, published in April 2014, projected US adults would spend 5 hours 46 minutes a day with digital media in 2014 against 4 hours 28 minutes with television, with mobile specifically identified as the driver of the shift rather than desktop internet use.
What makes 2013 a genuine crossover rather than a gradual drift is that it is a single, datable point where the lines on the chart cross. Before it, television commanded more daily time than digital. After it, digital held the lead and never gave it back. Media plans built on the assumption that television was still the default attention surface stopped matching reality at that point, whether or not the businesses running those plans noticed.
The second crossover: streaming over linear, 2025
Twelve years later, the same kind of event happened again, this time inside television itself. Nielsen's The Gauge, the industry-standard monthly panel for US TV-usage currency, recorded streaming reaching 44.8 percent of US TV usage in May 2025, the first time it exceeded broadcast and cable combined at 44.2 percent. By December 2025, streaming had climbed further to a new all-time high of 47.5 percent.
This crossover is narrower in scope than 2013, it is a shift within television viewing rather than across all media, but it is measured by the same kind of rigorous, panel-based methodology, and it marks the same kind of clean before-and-after line. Before May 2025, most television viewing in the US ran through broadcast and cable. After it, streaming held the larger share.
What connects the two events, twelve years apart
Both crossovers share three structural features worth naming directly. Each was measured by an established, methodologically disclosed panel rather than a single self-interested source, eMarketer's industry compilation for 2013, Nielsen's Gauge for 2025. Each has a specific date or narrow window attached to it, not a vague "sometime in the mid-2010s" or "recently." And each represents attention moving to a genuinely new place, not simply redistributing within an old one.
The twelve-year gap between them is itself informative. It suggests these regime-change events are not one-off anomalies but a recurring pattern in how attention reorganizes: a new surface builds quietly for years, then crosses a threshold that gets captured by whichever measurement panel happens to be tracking the right numbers at the right time.
The cost of reacting late
Neither crossover was hidden. eMarketer's 2013 figures and Nielsen's 2025 figures were both published, both covered by industry trade press, and both available to any advertiser paying attention. The businesses that reallocated spend toward digital in the years immediately following 2013, and toward connected TV and streaming inventory following the 2025 shift, moved with the data. The businesses that treated each shift as a trend to watch rather than a fact to act on kept buying inventory against a shrinking share of attention while paying for the illusion that nothing had changed.
This is measurable in the advertising market itself. eMarketer forecasts US connected-TV ad spend at approximately 37.95 billion dollars in 2026, up 14.5 percent year over year, and projects it will overtake linear TV ad spend by 2028. Ad budgets are already following the second crossover the same way they eventually followed the first. The question for any individual advertiser is not whether the reallocation is happening, the panels have already answered that, but whether they are moving with it or after it.
What the two crossovers do not prove
The two events are comparable in kind, both are dated, panel-measured points where a newer surface overtook an older incumbent, but they are not measuring the same thing at the same scale, and treating them as directly equivalent would overstate the case. The 2013 crossover compared total daily time across all of digital media against all of television, an entire-category comparison. The 2025 crossover compared streaming's share against broadcast and cable's combined share within television specifically, a narrower, within-category comparison.
Neither figure says anything definitive about a comparably sized shift now underway in AI-answer search, which is measured by different panels using different methodologies again. The value of the two crossovers is as a documented pattern, attention regimes do flip, on datable timelines, and the flip tends to be durable once it happens, not as a precise forecast of when or by how much the next one will land.
Why "regime change" is the right word, not "trend"
A trend implies something gradual, reversible, worth waiting on for more data before committing. A regime change implies a threshold has been crossed and the prior baseline no longer applies. Both 2013 and 2025 fit the second description better than the first: in both cases, once the crossover happened, the new leading surface did not lose its lead. Digital did not cede its position back to television after 2013. Streaming has continued climbing past its May 2025 crossover point through the rest of that year.
This is the basis for treating today's AI-answer shift as a live regime-change candidate rather than a hype cycle to wait out. Two prior shifts of this kind, independently measured, both proved durable rather than temporary once they crossed their threshold.
The evidence
Key findings, with their sources
-
In 2013, US adults' daily time with digital media first exceeded daily TV time; eMarketer projected 5h46m digital versus 4h28m TV for 2014, with mobile identified as the driver.
established eMarketer, "Mobile Continues to Steal Share of US Adults' Daily Time Spent with Media," April 2014.
-
Streaming reached 44.8% of US TV usage in May 2025, first exceeding broadcast and cable combined (44.2%), then reached a new high of 47.5% by December 2025.
established Nielsen, "The Gauge," news releases, 2025 to 2026.
-
US connected-TV ad spend is forecast at approximately $37.95 billion in 2026 (+14.5% year over year), projected to overtake linear TV ad spend by 2028.
established eMarketer, December 2025 CTV forecast.
Reference
Glossary
- The Gauge
- Nielsen's monthly, panel-based US television-usage report, treated as the industry-standard currency for measuring the split between broadcast, cable, and streaming.
- Linear TV
- Television consumed through a scheduled broadcast or cable feed, as opposed to on-demand streaming.
- Regime change (media consumption)
- A dated, measured point after which a new leading attention surface does not cede its position back, distinct from a reversible or gradual trend.
Straight answers
Frequently asked questions
When did digital media overtake TV in the US?
2013, based on eMarketer's industry-standard reporting, which projected US adults would spend more daily time with digital media (5h46m) than television (4h28m) for 2014, with mobile as the specific driver.
When did streaming overtake linear TV?
May 2025, according to Nielsen's The Gauge, when streaming reached 44.8% of US TV usage against 44.2% for broadcast and cable combined. Streaming continued climbing to a new high of 47.5% by December 2025.
Is linear TV going away?
The data here does not support that claim. It shows streaming has become the larger share of TV viewing, not that linear TV has disappeared. Broadcast and cable combined still commanded 44.2% of usage at the crossover point.
How fast is CTV ad spend growing?
eMarketer forecasts US connected-TV ad spend at roughly $37.95 billion in 2026, up 14.5% year over year, and projects it will overtake linear TV ad spend by 2028.
Provenance
Sources
- eMarketer, "Mobile Continues to Steal Share of US Adults' Daily Time Spent with Media," April 2014 (established)emarketer.com
- Nielsen, "The Gauge," news releases, 2025 to 2026, nielsen.com/data-center/the-gauge (established)
- eMarketer, December 2025 CTV ad-spend forecast (established)
Every figure above is attributed to a real, dated source and tagged with its evidence tier. Where a claim could not be verified to a primary source, it is not stated as fact.