Ad-supported streaming: the new free TV economy

Free television was supposed to be over. The cord-cutting wave of the mid-2010s was widely interpreted as the end of ad-supported broadcast — the moment when consumers, finally given the choice, voted with their wallets for paid streaming and ad-free viewing. By 2020, every major media company had launched a subscription service, and the working assumption inside the industry was that the future of television would be measured in monthly recurring revenue. The assumption was wrong. The fastest-growing layer of the connected-TV stack in 2025 and 2026 is the layer where the consumer pays nothing.

The free-tier resurgence has a name in the trade press — FAST, for Free Ad-Supported Streaming TV — and a dollar figure attached: roughly $33 billion in US connected-TV ad spend in 2025, projected to reach $38 billion in 2026 and $46.9 billion by 2028, the year ad-supported streaming is forecast to overtake linear television advertising for the first time in the medium's history.^[1] The growth is not a recovery of an old model. It is the construction of a new one, on top of infrastructure that did not exist five years ago. The audience, meanwhile, has its own word for the entire phenomenon — a word documented in our analysis of what “online TV” means in 2026 consumer search data.

Three things that changed

The free-TV economy did not return for one reason. It returned for three, in close succession.

First, smart-TV penetration crossed a threshold. By 2026, more than 75% of US households were projected to use connected TV regularly, with smart TVs themselves accounting for the majority of that usage by 2027 — meaning ad-supported channels could reach audiences without requiring those audiences to plug in any external hardware.^[2] The free service no longer competes against the cable box. It competes against the home screen of the device the cable box used to plug into.

Second, subscription fatigue arrived. Households that had stacked four or five paid streaming services through the early 2020s began consolidating, churning, and — in increasing numbers — adding free ad-supported tiers to fill the gaps left by the cancellations. Industry tracking through 2025 placed the number of streaming services projected to clear $1 billion in advertising revenue in 2026 at nine, up from two in 2020. Several of those services were ad-supported tiers of products that had launched as ad-free.^[3]

Third, ad-tech caught up. Programmatic CTV inventory — bought via the same demand-side platforms used for display and search — became the default purchase mechanism for performance advertisers between 2022 and 2025. Interactive ad formats arrived shortly after, with measured lifts of +71 seconds in viewer time, +36% in brand recall, and +13% in foot traffic for retail advertisers running interactive CTV creative; by 2026 around 41.8% of US marketers reported using interactive ads on connected TV.^[4]

The structure of the carriage layer

FAST distribution is concentrated. Most of the audience for free ad-supported streaming flows through a small number of aggregator platforms — typically operated by major media companies, smart-television manufacturers, or device makers — each of which curates a grid of hundreds of linear streams. Some streams are operated by the platform directly; others are operated by third-party programmers who effectively rent channel slots in exchange for a share of advertising revenue.^[5]

The aggregator layer in 2026 is structurally similar to the cable carriage layer of the 1990s. A small number of distribution gatekeepers curate access to a much larger universe of channels and programmers. The economic terms — what proportion of advertising inventory the carrier retains, what placement carriage carries with it — are negotiated bilaterally and rarely disclosed.

The economics of the platform layer are different from the economics of the programmer layer. Platforms monetise device sell-through, ad-inventory rev-share, and home-screen real-estate placement; programmers monetise the ad inventory inside their own channels minus the platform's cut. Both layers have substantial revenue at scale — but the platform layer captures a disproportionate share of the value, because the carriage decision happens at the platform level.

Why the free model wins where subscription doesn't

Free is not a price point. It is a removal of friction. The household that will not start a subscription will start a stream — and once the stream is started, the audience economics are identical.

The structural advantage of the free model is that it removes the credit-card decision from the acquisition funnel. Every paid streaming service, no matter how compelling its content, has to convert a household that has already churned out of two or three other services this year. The free service has to convert a click. The conversion-rate gap is enormous, and it compounds over time as the household's mental list of “services I won't bother with” grows longer. The mathematics of why a friction-free acquisition channel commands a structural premium are treated in our companion piece on how buyers value direct-navigation traffic in 2026.

This does not make the free model superior on every dimension. Per-viewer ad-revenue economics remain materially below per-viewer subscription economics for premium content, and a free service can only sustain content investment to the level its ad inventory will support. But for the marginal viewer — the household that would not have subscribed at any price — the free service is the only economic option that captures any value at all.

Where the next decade goes

Three forecasts, each with reasonable confidence intervals. First, the FAST share of total CTV viewing is likely to keep growing through at least 2028, before plateauing as the ad-load ceiling on free content tightens. Second, the consolidation of FAST platforms is likely to continue — there are too many aggregators chasing the same channels, and the smart-TV manufacturers in particular are advantaged by their device-side distribution control. Third, ad-supported streaming will, by the end of the decade, simply be television for most viewers, in the same way that cable became television in the 1980s and 1990s. The acronyms will fall away. The audience category vocabulary will not.

What this is not

This is a description of an industry trend, not an investment recommendation. Specific platforms named are reported on for journalistic reasons, and inclusion does not imply any current or anticipated commercial relationship.