What the category covers

AVOD stands for advertising-based video on demand. It describes any on-demand streaming video that the viewer selects from a catalogue and that is funded by advertising rather than by a monthly subscription fee or a per-title payment. The commercial bargain is simple: the content costs nothing or costs less, and the viewer sits through commercial breaks that the service sells.

Three quite different businesses sit inside the label. Free services such as Tubi, Pluto TV and the Roku Channel carry no subscription at all. Ad-supported tiers of paid services, sometimes called SAVOD, charge a reduced fee and still run advertising; Netflix, Disney+, Max, Peacock and Amazon Prime Video all operate one. Broadcaster catalogues, covered separately as BVOD, add a third group with its own trading conventions. Nielsen began splitting the category formally in its 2026 upfront guide, published 12 March 2026, separating scheduled FAST channels from on-demand AVOD for the first time in its demographic reporting.

The distinction that matters technically is the on-demand request. Because playback starts when the viewer presses play rather than at a fixed hour, every break is assembled at the moment of viewing. There is no pre-sold 30-second slot waiting in a schedule. Each break is a decision made in a few hundred milliseconds by an ad server responding to a request from the player or the video packager.

How the inventory is traded

A typical AVOD session works in stages. The player requests the stream. The manifest identifies cue points, the moments where advertising may be inserted, and VMAP or an equivalent schedule states how many breaks exist and where. At each cue point the ad server receives a request carrying the content metadata, device type, an advertising identifier, geography and the break specification, then returns creatives wrapped in VAST, the Video Ad Serving Template that tells the player what to fetch and which events to report.

Most premium AVOD stitches those creatives into the video stream itself using SSAI, or server-side ad insertion, rather than handing them to the player separately. Stitching removes the buffering gap between programme and commercial, defeats most client-side blockers and works on television hardware that cannot run a full ad software development kit. It also strips out a good deal of signal, which becomes a measurement problem discussed below.

Breaks are sold as pods rather than as single spots. A pod is a sequence of slots inside one break, and the fields that describe it entered the programmatic protocol with OpenRTB 2.6, released for public comment by IAB Tech Lab on 9 December 2021 and finalised the following year. Podid groups slots belonging to the same break. Podseq marks whether the break is first, mid-programme or last. Slotinpod identifies position inside the break, the streaming equivalent of a first-in-break premium. Poddur states the total seconds available, maxseq the number of slots, and mincpmpersec sets a floor priced per second, so a rate of two dollars implies a 30 dollar floor for a 15-second ad and 60 dollars for a 30-second one.

According to Index Exchange, the older method of requesting each slot separately produced 24 calls to three supply-side platforms for one eight-slot break; pod bidding reduces that to three. On the sell side the decisioning usually runs through Google Ad Manager, FreeWheel, Magnite's SpringServe or Publica. On the buy side the spend arrives through upfrontcommitments, programmatic guaranteed deals, private marketplaces and, for a minority of volume, the open exchange.

Ad load is the lever that sets supply. Services generally run four to six minutes of advertising per hour against the 14 to 18 minutes common on linear television, concentrated in mid-roll positions. Research from ShowHeroes and Omnicom Media Netherlands, published 24 March 2026, put viewer tolerance at about four minutes per hour.

Origin and evolution

Ad-funded on-demand video predates the acronym. Sony relaunched Grouper as Crackle in 2007, and NBC Universal and News Corporation opened Hulu to a private beta in October that year before a public launch on 12 March 2008. Hulu established the template that the industry still uses: professionally produced episodes, a light commercial load, and measurement borrowed from television rather than from display.

Standards followed the money. VAST 1.0 arrived in 2008. The IAB Video Suite of April 2012 paired VAST 3.0 with VMAP 1.0, giving publishers a way to describe break structure independently of the creatives that would fill it. VAST 4.0 landed in 2016 and separated the media file from the interactive layer, an important step for television hardware. Version 4.2 followed in June 2019 and version 4.3 in December 2022.

The supply side changed shape twice. Pluto TV and Tubi, both founded in 2014, proved that free catalogues could scale without a broadcaster parent. Then the subscription giants reversed course. Netflix launched Basic with Ads in the United States on 3 November 2022 at 6.99 dollars a month, undercutting Disney+, which introduced its own ad tier on 8 December 2022 at 7.99 dollars. Amazon went furthest on 29 January 2024 by switching advertising on for existing Prime Video subscribers by default and charging extra to remove it, converting a subscription base into an advertising audience overnight.

Why it matters to the marketing community

Scale is the short answer. Nielsen's Ad Supported Gauge for the second quarter of 2026 found advertising-funded television at 71.5% of total United States viewing, with streaming at 48.2% of that total. Among adults 18 to 49, streaming took 66.7% of ad-supported television time, and non-FAST AVOD accounted for 81.1% of streaming time in that group against 19% for FAST.

The Video Advertising Bureau's twelfth annual streaming study, released 31 March 2026, counted 209.4 million ad-supported viewers in the United States, equal to 62% of the population. The same report found 46% of premium subscription accounts now sit on an ad-supported plan, up from 33% in 2023, and projected 38 billion dollars of connected television ad spend in 2026, or 43% of all television advertising.

European figures point the same way with a different mix. The ShowHeroes survey of 4,377 consumers across six markets recorded 27% household adoption of FAST channels, rising to 37% in the United Kingdom and Italy, and found 47% of German FAST viewers unreachable through subscription ad-supported platforms.

Limitations and disputes

The most contested number in AVOD is completion. Non-skippable formats on television screens finish almost automatically, which makes the metric a description of the ad unit rather than of attention. Paid media consultant Austin Lake argued in June 2026 that a 98% completion rate is structural, not earned, and that optimising towards it pushes budget into cheap inventory. Vlad Chubakov of Delve Deeper put the figure closer to 99.99%. DoubleVerify found 34 of every 100 monitored impressions did not land in streaming television content at all, and counted a 140% rise in connected television fraud schemes in the first quarter of 2026 against the same period a year earlier.

Trade body promotion of the metric has drawn its own criticism. A VAB email of 28 August 2026 headlined a 94% completion rate that originated in FreeWheel research from September 2023, alongside a 97% figure measured on a different basis.

Viewer behaviour undercuts the delivery numbers further. A survey of 1,000 United States adults published 22 April 2026 found only 24% pay attention to streaming advertising, while 74% turn to a second screen and 38% leave the room; 52% had considered cancelling a service because of its ad load, and 67% preferred a single long pre-roll to scattered breaks.

Two structural problems remain unresolved. Frequency capping leaks across supply paths, because the same household can be bought through a direct deal, an exchange and a reseller without any of the three recognising the duplication, and SSAI removes some of the identifiers that would catch it. Measurement bases also move the answer: Nielsen's own data shifts streaming from 48.2% to 44.4% of ad-supported viewing when the base changes from total persons to adults 18 and over.

Not the same as

SVOD is subscription video on demand, funded entirely by recurring fees. An SVOD service with an ad tier operates AVOD inventory inside an SVOD business.

TVOD is transactional video on demand, a rental or purchase of a single title, with no advertising at all.

FAST is free ad-supported streaming television: scheduled linear channels delivered over the internet. The funding model matches AVOD, the consumption model does not, because the viewer joins a running channel rather than choosing a title.

CTV is a device classification covering smart televisions and streaming sticks. AVOD is consumed on connected televisions, phones and desktops alike, so the two terms describe different axes.

Recent developments

Netflix told buyers at its upfront on 13 May 2026 that its ad-supported plan reaches 250 million monthly active viewers, that advertising revenue should approach 3 billion dollars in 2026 against roughly 1.5 billion in 2025, and that programmatic buying is nearing half of non-live inventory. Fifteen further ad markets open in 2027. Earlier, on 4 March 2026, Netflix added Amazon shopping audiences, Yahoo signals and its own conversion API for server-to-server attribution.

Amazon reported 21.3 billion dollars of advertising revenue in the fourth quarter of 2025 and an average ad-supported Prime Video audience of 315 million viewers across 16 countries. On 10 April 2026 it raised the price of removing those ads from 2.99 to 4.99 dollars and rebranded the ad-free option as Prime Video Ultra.

Timeline

  • 2007: Sony relaunches Grouper as Crackle; Hulu opens a private beta in October
  • 12 March 2008: Hulu launches publicly in the United States
  • 2008: IAB publishes VAST 1.0
  • April 2012: IAB Video Suite pairs VAST 3.0 with VMAP 1.0
  • 2014: Pluto TV and Tubi founded
  • 2016: VAST 4.0 separates media files from the interactive layer
  • June 2019: VAST 4.2 released
  • 9 December 2021: IAB Tech Lab opens OpenRTB 2.6, with pod bidding, for public comment
  • 3 November 2022: Netflix launches Basic with Ads at 6.99 dollars
  • 8 December 2022: Disney+ launches its ad tier at 7.99 dollars
  • December 2022: VAST 4.3 released
  • 29 January 2024: Amazon switches advertising on by default for Prime Video subscribers
  • 4 March 2026: Netflix adds Amazon audiences, Yahoo signals and a conversion API
  • 12 March 2026: Nielsen splits FAST from AVOD in its upfront guide
  • 24 March 2026: ShowHeroes reports 27% European FAST adoption
  • 31 March 2026: VAB counts 209.4 million United States ad-supported viewers
  • 10 April 2026: Prime Video Ultra replaces the 2.99 dollar ad-free option
  • 13 May 2026: Netflix reports 250 million monthly ad-supported viewers
  • 25 August 2026: Nielsen puts ad-supported viewing at 71.5% of the United States total

Summary

Who: Streaming services that sell advertising, from free catalogues such as Tubi and Pluto TV to the ad tiers of Netflix, Disney+, Max, Peacock and Amazon Prime Video, together with the ad servers, supply-side platforms and demand-side platforms that trade the breaks.

What: On-demand streaming video funded by advertising instead of subscription fees, assembled break by break at the moment of viewing and traded as pods rather than single spots.

When: Commercially established from Hulu's 2008 launch, standardised through VAST and VMAP between 2008 and 2022, and reshaped by the subscription services adding ad tiers between November 2022 and January 2024.

Where: Consumed mainly on connected televisions but also on phones, tablets and desktops; largest in the United States, where ad-supported viewing reached 71.5% of total television time in the second quarter of 2026, with rapid growth across the United Kingdom, Germany, France, Italy, Spain and the Netherlands.

Why: Subscription growth slowed and advertising reopened a second revenue line, giving buyers television-style reach with digital targeting and giving viewers a cheaper way in. The trade-off is measurement that flatters delivery and attention data that does not.