Streaming Won, but There Are Two Kings

In December 2025, streaming hit an all-time high, taking 47.5% of US TV viewing. It's easy to read this as "a total win for OTT." Yet across all of US TV — in Nielsen's distributor tally that combines linear broadcast and streaming — the single distributor that grabbed the most screen time was not Netflix but YouTube. Come to Korea and the gap widens. The time one person spends on short-form apps in a month is about seven times the combined total of the five major OTT services (48.7 hours versus 7.2 hours). So who, really, is the "king of media"?

'Strength' Has Two Kings — Revenue Is Games, Attention Is Creators

We have to start by taking apart the phrase "media is strong." Strength is not a single metric but a word where revenue, attention (viewing time), and the movement of ad money overlap — and those axes don't point to a single king. Split honestly, there are two.

By absolute revenue, the champion is games. In 2025, global game revenue was $188.8–197 billion, larger than film and music combined. Games are also #1 on the "most firmly measured" axis — unlike other domains where estimates split by orders of magnitude across firms, absolute revenue here is sharp, and games rank #1 in Korean content exports too. Growth is still positive, as well (it's merely in a slowdown phase). When this piece says "creator strength," it is not pulling games down.

The king of the other axis — attention, and the ad momentum chasing that attention — is neither OTT nor Netflix. It is creator video, centered on YouTube. The YouTube viewing Nielsen counts is filled mostly by creators and UGC, not studio productions, so "#1 platform" is the same as "#1 in creator video" (who captures that value is a separate question, which we'll get to shortly). The creator economy is cited as the steepest-growing segment in media (absolute market-size estimates split down to the order of magnitude across firms, so we read only the direction). Growth is not a separate axis but the shadow of this attention shift.

Attention: YouTube Leads OTT

What we can say most firmly is the attention axis. In Nielsen's tally, YouTube held the #1 spot among single US TV distributors for a year starting July 2025 (13.4–13.9%). Disney at 10.3%, Paramount at 7.9%, and Netflix at 7.8% follow. One thing to note. This share is measured not "within streaming" but against all TV, linear broadcast included, and Disney's and Paramount's figures fold in each company's cable and broadcast networks (ESPN, CBS, and the like) wholesale. YouTube, despite being streaming-only, comes in ahead of those totals. And because this figure is measured on the TV screen, mobile viewing is undercounted — YouTube's lead may, if anything, be a conservatively estimated value.

Korea is more dramatic still. Short-form app usage time is about seven times that of OTT, and YouTube is the most-used app, used by 91.5% of smartphone users (MAU 46.82 million). (The other branch — where attention is outsourced to short-form and creators — I covered in the outsourcing of choice to short-form.)

The Money Follows Attention, but the Platform Holds the Money

Advertising growing at three times the pace of consumer spending is the engine of media's current strength. And that money follows where attention has gone. In 2025, US social ad spend was $83.3 billion, about 30% of digital advertising (within the "display-type" segment that excludes search, it crosses half), and the fastest-growing ad category PwC names is also social and short-form video.

Here is the non-obvious part. The king of attention and advertising is creator "content," but the one that books that value into the ledger is not the creator. The accounts the $83.3 billion flows into are named YouTube (Google), Instagram (Meta), and TikTok (ByteDance) — the three platforms that hold the rails. Creators generate the content, and the platform oligopoly captures the value on top of it. (This is analysis, not a verdict — creator revenue shares vary by platform and contract. What is clear is who holds the chokepoint of the ad transaction.)

The same capture works on the premium-video side too. Connected-TV advertising grew about 16% in a year to reach $33.4 billion, and ad-supported free streaming (FAST) viewing jumped 43–55%. The apparent contradiction — "FAST is rising while commissions for scripted drama are being cut" — resolves into one explanation: new premium commissions are being trimmed, while catalog already produced appears to be re-distributed cheaply through FAST channels. The agent of monetization, in effect, shifts from the studio to the platform.

Two Strong Genres — Creator Video and Live Sports

Narrow it to genre and strength runs in two branches. One is creator video. Even podcasting, once the domain of audio, has seen its most-frequently-used platform shift to YouTube (44% of US listeners · 1 billion monthly viewers). In nonfiction, true crime is growing fast. The other is live sports. In 2025, 96 of the top 100 most-watched US content titles were sports (narrow to the top 50 and it's 46 of 50).

People commonly say "these two are safe because they're hard to copy or synthesize." Only half right. AI imitates a creator's "content" as much as you like. What withstands synthesis is not the content itself but liveness (the match has to happen now) and a relationship of trust with a particular person. The reason sports is strong and the reason creators are strong are different from each other.

Korea: Games Hold It Up; the World Wants the Drama, but Commissions Get Cut

Korea is not a condensed version of the global pattern but an amplified one. Games hold up the absolute scale — of Korea's $14.07 billion in content exports in 2024, games were 60.4% ($8.5 billion), and music and broadcast video combined still fall short of half of games. It's a number that runs head-on against the conventional wisdom that "K-drama and K-pop are the export mainstay."

The real fissure is in drama. In Netflix's global viewing, Korean content ranks #1 outside the US, at 12.1 billion hours (a third-party estimate by Omdia and Digital-i). Even with the world demanding it this much, domestic scripted commissions fell 39% from the first half of 2023 to the first half of 2025 (global streamer commissions -43%, a single estimate from Ampere Analysis). There's no single culprit for the commissioning drop — (1) a base effect as the 2021–22 over-commissioning returned to normal, (2) a surge in production costs and talent fees, and (3) Netflix's full-buyout IP model appear to overlap.

But the one line that remains even after stripping all three away is the crux. Global demand is verified (12.1 billion hours), yet the capital to take in that demand does not return home. The platform that buys the IP outright captures the margin, and only the cost is left in the producing country. Creator content and K-drama point in a similar direction but differ in degree — K-drama hands over the IP wholesale, so its upside is near zero, whereas creators split the revenue and, owning their channel and audience, collect a toll on the chokepoint. In both cases the producer is the one who makes it; the one who holds more is the side that owns the rails. (Webtoons, at least, are separately strong as an IP source — 2024 industry size of 2.2856 trillion won, +4.4%.)

Set all this against conventional wisdom and two things flip.

So What Comes Next

From here on it's inference, not measurement. I write with the strength of conviction turned down.

That money follows attention is backed by data. One step further — "attention follows trust" — is an assumption (there are no figures to back it). If this assumption holds, the winners are the camps that hold attention and trust together. Live sports and creators — as we saw, the sides where it's not the content but liveness and trust in a person that withstand synthesis. As for the revenue model, the trend of ad-supported formats (FAST, AVOD) hardening into the default is relatively clear. The FAST market grows from $12.28 billion in 2025 to $14.88 billion in 2026, and a third of the US population watches FAST at least once a month.

There's a slower, more hypothetical zone too. The more AI-synthesized output floods in, the more provenance — the production history that guarantees "what is real and who made it" — and human curation become a premium. Regulation has dropped an anchor — the EU AI Act's Article 50 mandates, from August 2, 2026, both labeling of generative-AI output and machine-readable watermarking together (violations up to €15 million or 3% of revenue; December 2 is merely a provisional grace period limited to AI already on the market before then). But even if the effective date is settled, enforcement and market response are not yet verified. On the other side of that, I expect commodity scripted and indiscriminate synthetic output to enter deflation.

This forecast can be tested by a few signals. Whether YouTube holds the 13% range in US TV distribution; whether sports holds the 80% line of the most-watched top 100; whether FAST growth doesn't break; whether the AI Act's labeling is actually enforced. If signals run the other way, the forecast has to be folded.

Extrapolate the single phrase "media is strong" straight onto the trend and you'll be wrong. The king of revenue is games, and the king of attention is creators — two different kings. And the one that books the value of that creator content is not the side that made the content but the platform that holds the rails. Ask which screen took the most people's time last night, and the answer tilts ever more toward YouTube rather than Netflix — together with the name of the account into which that screen's ad money flows.