FAST Channels Advertising: Platforms, Inventory, Ad Loads and Revenue Models

FAST stands for free ad-supported streaming television. Roku's own filing defines it as free, ad-supported linear streaming that excludes on-demand content (Roku 10-Q, Q3 2026). The viewer opens a guide, picks a channel and joins whatever is playing. Nobody chooses an episode or pays a fee.

That makes FAST the streaming version of a cable channel. AVOD, or ad-supported video on demand, is different: the viewer picks a title from a library and ads play around it. Most services now mix the two. Tubi offers more than 340 linear channels next to a library of over 350,000 movies and episodes (Fox 10-K, FY2026). The Roku Channel carries 550+ live linear channels in the US and more than 80,000 free titles (company figures) (Roku, 16 July 2026).

For a buyer, the difference matters in two places. Linear breaks arrive on a schedule, so supply spikes at fixed times and many viewers see the same break together. And many channels are built around one show or brand. Roku's Q3 2026 letter, for example, mentions a Law & Order FAST channel recently added by NBCUniversal (Roku, 2 November 2026). Channel-level targeting therefore works much like buying a TV network. The connected TV (CTV) device is the same in both cases.

A simple example shows the buy. A home-improvement retailer wants weekend reach among homeowners. It could buy a home-renovation FAST channel by name through a private deal, then add household audience targeting on top. A cable buy would have bought the whole network schedule instead.

FAST platforms come in three kinds. TV makers build a service into the set, such as Samsung, LG and VIZIO. Operating-system owners do the same on their devices, such as Roku and Amazon. Media companies run their own services that appear on every device, such as Tubi and Pluto TV. The table uses each owner's filings, releases and ad sales pages, checked on 9 November 2026.

The figures are not comparable with one another. Samsung counts monthly users worldwide, Nielsen measures share of US TV time, and the channel counts come from each owner's own pages. Use the table to see who owns what and where the door is, not to rank size.

Three changes in 2026 reshaped the ownership column. Fox now owns Tubi and Roku and placed both in a new Fox Digital Platforms division (Fox, 6 November 2026). Pluto TV's parent took over Warner Bros. Discovery and renamed itself Skydance Corporation (Skydance 8-K, 6 October 2026). And Walmart, which owns VIZIO, completed its purchase of the self-serve streaming TV platform Vibe.co on 4 August 2026 (Walmart via Vibe.co, 4 August 2026). For buyers this means fewer sales doors and more first-party data behind each one.

A FAST channel starts as a playout: a scheduled stream assembled from a content library. The channel owner marks each break with a cue. The common format is SCTE 35, which its standards body calls the core signalling standard for advertising breaks (SCTE 35-1 2023r2, accessed 9 November 2026). When a viewer's stream reaches a cue, a server-side ad insertion (SSAI) service asks an ad server what to play and stitches the ads into the video.

1. The viewer tunes in. The TV app requests the channel's stream from the platform.
2. A cue arrives. The stream signals a break of, say, two minutes.
3. The ad server decides. Whoever controls that break, the platform or the channel owner, runs a decision across direct campaigns and programmatic demand.
4. Programmatic demand bids. An SSP sends an OpenRTB request describing the pod, the channel and the device to DSPs, which bid for one or more slots.
5. The stitcher inserts the ads. Ads arrive as VAST responses, are transcoded to match the stream, and play without a gap.
6. Beacons fire. The SSAI service or the player reports impressions back to the ad server and to measurement vendors.

The engineering details of step 5, including how beacons and device data pass through the stitcher, are covered in SSAI and CTV ad serving. Choosing the ad server for step 3 is the subject of our comparison of CTV ad server options.

OpenRTB 2.6 added the fields a FAST break needs. The content.livestream value 1 marks a scheduled broadcast, and new Channel and Network objects name the channel and its parent company (OpenRTB 2.6). Pod fields describe the break, and inventorypartnerdomain tells crawlers which content owner may sell part of an app's supply. The latest release of the specification is 2.6-202606, published on 11 June 2026 (IAB Tech Lab on GitHub).

The example below is an illustrative excerpt for one slot in a two-minute break. Identifiers and domains are placeholders; field names and values follow OpenRTB 2.6 and AdCOM 1.0.

The Channel and Network objects are optional in the specification, so ask each SSP which fields it populates. The planned OpenRTB bid request builder will include a CTV template, and OpenRTB 2.6 explained walks through the rest of the request.

FAST services advertise lighter breaks than cable, and that promise is part of the pitch to viewers. There is no audited, cross-platform ad-load figure. What exists are company statements and vendor compilations, so the table labels each by type.

Two cautions apply. The vendor figures date from August 2023 and may have moved since. And a stated load describes the size of the break, not how many paid ads actually ran in it. Buyers who care about clutter should ask for the average pod length and ad count actually delivered on the channels in their plan.

Why does ad load matter to a buyer? A shorter pod means each ad has fewer neighbours, which can help recall. It also means fewer slots to sell, so platforms with lighter loads need higher prices or more viewing hours to earn the same revenue.

Every FAST channel runs under a carriage agreement between the content owner and the platform. The ad money is divided in one of two ways. Amagi, a FAST technology vendor that runs a marketplace for these deals, calls both "industry-standard" models (Amagi, accessed 9 November 2026).

Sources: Amagi, accessed 9 Nov 2026; broadpeak.io, 18 Aug 2023.

Public filings show the inventory side of the bargain. Roku says its ad inventory includes video inventory in The Roku Channel and ad inventory it obtains as consideration from distribution agreements with content partners (Roku 10-Q, Q3 2026). In plain terms, apps and channels that want carriage give the platform some of their ad minutes. The percentages in individual deals are not disclosed.

Inventory splits also need supply-chain plumbing. When a content owner sells minutes inside a platform's app, the app's ads.txt file lists the content owner as an inventory partner, and the content owner publishes its own authorised sellers. OpenRTB 2.6 carries that domain in inventorypartnerdomain and points to ads.txt 1.1 for the rules (OpenRTB 2.6). Our guide to the SupplyChain object, ads.txt and sellers.json shows how buyers check those files.

A worked example. Take a classic-TV channel with 50,000 viewing hours a month on one platform. At the 8 ad minutes an hour that Roku reports for its channel, those hours hold 400,000 ad minutes, or 800,000 thirty-second slots (Roku, 16 July 2026). Every other number below is illustrative.

1. Revenue share. The platform fills 60% of the slots at a $12 CPM. That is 480,000 impressions and $5,760 of revenue. At a 50% share, the channel owner receives $2,880.
2. Inventory split. The channel owner keeps 2 of the 8 minutes, or 200,000 slots, and sells them itself. To match $2,880, it must earn $14.40 per thousand slots, for example 80% fill at an $18 CPM.

The arithmetic explains the market's habits. A broadcaster with strong direct sales can beat a revenue share on its own minutes. A small library owner usually cannot fill its slots at that rate, so it takes the share and the platform's sales team. The broadpeak.io analysis adds a data point: channels push for inventory deals partly to see more about their own audience (broadpeak.io, 18 August 2023).

There are four main doors, and most large buyers use more than one.

1. The platform's own sales and DSP. Amazon sells Fire TV Channels through its DSP. On 29 September 2026 it folded the DSP into Amazon Ads Agent, where programmatic campaigns are now called DVA+ (Amazon Ads, accessed 9 November 2026; PPC Land, 29 October 2026). Our Amazon DSP profile covers the buying side.
2. Open DSP integrations. Roku says it integrates with every major DSP and SSP and planned to expand its DV360 integration in the coming months (company statement) (Roku shareholder letter, 2 November 2026).
3. Retail data on TV supply. Walmart Connect opened VIZIO inventory to its audiences through Yahoo DSP, using Magnite's supply-side technology, in a closed proof of concept announced on 28 May 2026 (Walmart Connect, 28 May 2026). The Walmart Connect profile tracks that programme.
4. Channel owners' own supply. Under an inventory split, the content owner sells its minutes through its own ad server and SSP. Buyers reach these through deals with the network, often across several platforms at once.

A deal ID lets a buyer agree a price and a channel list in advance, then bid in real time. The differences between a private marketplace (PMP), a preferred deal and programmatic guaranteed are set out in programmatic guaranteed vs PMP.

Self-serve tools are spreading too. Walmart describes Vibe.co, which it now owns, as a self-service platform that has helped thousands of businesses buy connected TV (Walmart via Vibe.co, 4 August 2026). For a small brand, that is often the first route into FAST breaks.

Before buying, check the supply path. The same channel can appear on several platforms, each with its own sellers: Amagi alone lists 400+ channels and 50+ platforms in its marketplace (company figure) (Amagi, accessed 9 November 2026). Confirm the app in app-ads.txt, the seller in sellers.json and the channel name in the bid request. If the channel fields are empty, ask the seller why.

At platform level, measurement is maturing. Nielsen reports The Roku Channel separately in its Media Distributor Gauge, while Pluto TV sits inside Paramount's total (Nielsen, accessed 9 November 2026; Roku, 16 July 2026). Nielsen notes that the Gauge does not reflect its currency TV ratings.

TV makers add their own data. VIZIO says its Inscape ACR data comes from more than 28 million opt-in devices (company claim) (VIZIO Ads, accessed 9 November 2026). LG Ad Solutions and Samsung Ads both sell deterministic data from their sets (company descriptions) (LG Ad Solutions; Samsung Ads). That data can build a household graph for targeting and for matching exposure to outcomes.

The weak point is the channel. Because the channel and network fields are optional, a bid request can arrive without them. A buyer may know it reached a household on Samsung TV Plus without knowing which channel ran the ad. Our comparison of CTV measurement companies covers the vendors that verify CTV delivery and audiences.

1. List channels, not just platforms. Ask each seller for a channel list with genres and the share of your impressions each channel delivered.
2. Ask who sells each minute. Find out whether the platform or the channel owner holds the inventory you are buying, so you know whose data and rules apply.
3. Check pod length and position. Ask for average ads per pod and whether first-in-pod positions can be bought.
4. Insist on channel fields. Require content.channel and content.network in bid requests, or channel-level reporting in direct deals.
5. Confirm authorisation. Check app-ads.txt, sellers.json and any inventory partner domain before scaling spend.
6. Plan frequency across services. The same channel often runs on several FAST platforms, so cap frequency at household level.
7. Measure outcomes. Use ACR or retailer data where offered, and hold out a control group before judging results.

1. Fox Corporation, Form 8-K, exhibit 99.1, "FOX Corporation Completes Acquisition of Roku" (6 November 2026). https://www.sec.gov/Archives/edgar/data/1754301/000119312526451888/d141597dex991.htm
2. Roku, Inc., Form 8-K on completion of the merger with Fox Corporation (6 November 2026). https://www.sec.gov/Archives/edgar/data/1428439/000119312526451874/d181620d8k.htm
3. Roku, Inc., Q3 2026 Shareholder Letter, Form 8-K exhibit 99.1 (2 November 2026). https://www.sec.gov/Archives/edgar/data/1428439/000162828026068877/a04fy2026q3exhibit991.htm
4. Roku, Inc., Form 10-Q for the quarter ended 30 September 2026 (2 November 2026). https://www.sec.gov/Archives/edgar/data/1428439/000162828026068887/roku-20260930.htm
5. Roku Advertising, "Everything advertisers want to know about FAST TV, answered" (16 July 2026). https://advertising.roku.com/learn/resources/everything-advertisers-want-to-know-about-fast-tv-answered
6. Fox Corporation, Form 10-K for fiscal year ended 30 June 2026 (6 August 2026). https://www.sec.gov/Archives/edgar/data/1754301/000162828026053960/fox-20260630.htm
7. Tubi, corporate site "Advertise With Us" (accessed 9 November 2026). https://corporate.tubitv.com/
8. Nielsen, The Gauge and Media Distributor Gauge, August 2026 data (accessed 9 November 2026). https://www.nielsen.com/data-center/the-gauge/
9. Samsung Electronics, "From Hardware Roots to Global Media Platform: How Samsung TV Plus Reached 100 Million Users" (13 February 2026). https://news.samsung.com/global/from-hardware-roots-to-global-media-platform-how-samsung-tv-plus-reached-100-million-users
10. Samsung Ads, US home page (accessed 9 November 2026). https://www.samsung.com/us/business/samsungads/