A recoup rate is the proportion of advertising money leaving one medium that comes back to the same sellers somewhere else. In television it has a precise form: of every unit of currency that exits linear broadcast, what share lands on the broadcasters' own streaming inventory rather than leaving the sector. The remainder goes to Google, Meta or Amazon.

The measure exists because the defence of television economics rests on an assumption that migration is internal. Audiences move from the broadcast stream to the on-demand service, and budgets are supposed to follow into the same accounts at a higher price per thousand impressions. A recoup rate tests that with spend records rather than viewing figures, asking where the money went rather than the audience.

How the number is built

No ad server reports a recoup rate. It comes from spend-tracking firms working off agency billing data, the only vantage point from which a budget can be watched leaving one line and arriving in another. Guideline, the New York ad intelligence company led by Vince Mifsud, tracks roughly $200 billion of annual media investment across 65 countries and extended its SQADCosts pricing benchmarks from linear television into connected television, audio streaming and podcasting on April 30, 2026. Its client base spans the holding companies and large independents, which sets both the strength and the boundary of the figure.

The underlying technique is source-of-volume analysis, which reconstructs a channel's growth from the channels that funded it. A worked example published in March 2026 runs the arithmetic in the other direction: United States out-of-home advertising started 2024 at $3,664.64 million, and digital performance channels took $104.84 million out of that pool while television added $248.25 million and connected television added $19.40 million, carrying the 2025 total to $3,859.30 million. A recoup rate is one cut of the same ledger, filtered to a single seller group and a single origin channel. The numerator is money arriving on broadcaster video on demand; the denominator is money confirmed as having left linear. Neither side is standardised in the way the Media Rating Council accredits.

What the ratio currently shows

The figures entered public circulation through a strategic assessment published in July 2026 by Liberty Sky Advisors, the independent advisory firm run by Ian Whittaker, a former City equity research analyst. The report, titled A Risk Too Far, argues against switching off United Kingdom digital terrestrial television in 2034.

According to Guideline data relayed on the Media Unfiltered podcast by the firm's chief insights and analytics officer Sean Wright, roughly 15p of every pound leaving the UK linear market is recouped on broadcaster streaming, against around twenty-five cents in the dollar in the United States and twenty cents in Canada. Of the three markets, the United Kingdom recovers the least. The report labels the figures as pending confirmation of permitted use, a caveat worth carrying.

The price structure behind the number

Price explains the gap. On Guideline's dollar-denominated data, UK linear and pay-television equivalent inventory clears at $4 to $5 per thousand impressions across all adults, while UK broadcaster video on demand sits at roughly $22 to $25, a multiple of about five. YouTube buying prices sit near $5, inside the linear band.

An advertiser leaving linear is therefore invited to pay five times more to stay with the same seller, or roughly the same price to leave. Whittaker calls the resulting behaviour cheap-tier retention, and it inverts the broadcaster's logic: linear is the affordable tier of television, and a switch-off removes the one tier where broadcasters compete with the platforms on price. The premium that migrated off the linear rate card into streaming is what makes streaming hard to sell at volume.

The pull on the other side is documented. YouTube generated approximately $36.1 billion in advertising revenue in 2024 against a global television market of roughly $180 billion, with growth decelerating from 45.9 per cent in 2021 to around 12.5 per cent in 2025, which is why it needs television money. The pitch was made at Brandcast on May 13, 2026, with two-click connected television checkout and expanded retail data partnerships aimed at buyers weighing television budgets.

Origin and evolution

Budget-flow analysis long predates the phrase. What is new is applying it to a single seller group as a retention ratio, and the wording remains unsettled: no trade body defines a recoup rate, and no industry glossary carries it.

The policy context that produced the number is dated. United Kingdom multiplex licences were extended to 2034 under a 2021 order, and Ofcom published its early market report on television distribution on May 9, 2024. The government's green paper, Watch this Space: A new strategic direction for UK media, arrived on June 23, 2026, consulting on withdrawal in 2034 or an extension to 2044. The Liberty Sky report followed as an argument for the later option, with the recoup rate as the hinge of its advertising case.

Why it matters for media buyers

What holds linear revenue up, on the report's reading, is not a price premium but a classification. Linear, broadcaster streaming, digital video and connected television sit in separate lines of the agency trading framework, with separate teams and budgets, and Nielsen's 2026 Upfront Planning Guide quantified the resulting gap on March 12, 2026: streaming took 66.7 per cent of ad-supported television time among adults 18 to 49, while linear captured 67.5 per cent of television advertising spending. A dated end for broadcast collapses that silo, and once the lines merge the recoup rate decides how much of the pool broadcasters keep.

Holding companies have their own interest in that merger. Inventory bought as principal and resold at an undisclosed markup carries higher margin than inventory bought transparently as agent, and principal media has returned to the centre of holding company strategy. High-quality linear sits closest to the agent model; connected television and programmatic sit closest to the principal one. Concentration sharpens it: Omnicom completed its acquisition of Interpublic on November 26, 2025, and WPP pools client spending to negotiate with Google, Meta, Amazon and broadcasters as a single buyer.

The money at stake is specific. Liberty Sky estimates commercial public service media advertising revenue of about £3.9 billion in 2026, drifting to £3.64 billion by 2034 if terrestrial broadcast continues, against under £3.1 billion if it ends in 2034, a fall of roughly 16 per cent, with cumulative losses near £3.5 billion between 2028 and 2036. Steve King, former chief executive of Publicis Media, is quoted saying the broadcasters push hardest for switch-off and that it may prove the thing that kills their own premium product.

Limitations and disputes

The figure is an inference about destination rather than an observation of it. A budget that leaves linear and appears on YouTube in the same quarter has not necessarily moved, since source-of-volume models allocate flows by pattern. Coverage is also partial: agency-sourced panels see agency-placed money, leaving the self-serve layer invisible. Whittaker has himself argued that the advertising market is two markets and most forecasts see only one, a critique that applies to his own inputs.

The two parties disagree on where prices go next. Guideline holds that UK broadcaster streaming is underpriced against international peers and has room to grow. The Liberty Sky report argues the likelier path is convergence downward toward the YouTube band, and states that if Guideline is right, its own findings weaken.

Price direction is contested elsewhere too. Guideline data puts video on demand and connected television prices down 30 to 40 per cent over eighteen months, while DataBeat's May 2026 report recorded connected television at $4.68 per thousand impressions, down 25.8 per cent year on year as overall programmatic prices rose 34 per cent. Managed-service ranges in vendor research run far higher. Those numbers describe different markets, and a recoup rate built on one inherits the ambiguity.

Volume confounds it further. Ad-supported streaming reached 209.4 million United States viewers, with connected television spending projected at $38.0 billion in 2026. Falling prices against rising supply can still raise revenue, so a low recoup rate is not a revenue forecast.

Not the same as

A fill rate measures whether an ad request returned an advertisement. It concerns unsold inventory inside one seller, not budget moving between them.

A payback period, called a recoup day in application marketing, is the point at which a campaign has returned what was spent on it. That is about time and return on advertising spend, not about where a departing budget lands.

A recoupment rate in royalty accounting sets how fast revenue is credited against an advance before payments resume. The wording is close and the subject unrelated.

Incrementality asks whether advertising caused an outcome. A recoup rate asks only where the invoice went.

Recent developments

The consolidation running alongside the consultation changes who the ratio describes. Sky agreed terms to buy ITV's broadcast and streaming unit for £1.6 billion in June 2026, combining the two largest UK premium video sellers, with ITV retaining public service status to 2034 under the announced terms. Comcast then announced on June 29, 2026 that it intends to separate NBCUniversal and Sky through a tax-free spin-off, putting the buyer under restructuring of its own.

Sell-side cooperation runs the same way. Sky, Channel 4 and ITV announced a unified self-service television advertising marketplace on Universal Ads and FreeWheel technology on June 17, 2025, aimed at smaller advertisers. A single route across three sales houses lifts the recoup rate only if the prices behind it stay competitive with the platforms, which is the question the metric was built to expose.

Timeline

  • September 2021: United Kingdom digital terrestrial multiplex licences are extended to 2034
  • May 9, 2024: Ofcom publishes its early market report to government on the future of television distribution
  • June 17, 2025: Sky, Channel 4 and ITV announce a unified self-service television advertising marketplace
  • November 26, 2025: Omnicom completes its acquisition of Interpublic
  • March 12, 2026: Nielsen publishes its 2026 Upfront Planning Guide
  • April 30, 2026: Guideline extends SQADCosts Local pricing benchmarks into connected television, audio streaming and podcasting
  • May 13, 2026: YouTube uses Brandcast to pitch television buyers on connected television checkout and retail data
  • June 23, 2026: The United Kingdom government publishes Watch this Space, consulting on terrestrial switch-off in 2034 or 2044
  • June 24 to 26, 2026: Sky agrees terms to acquire ITV's broadcast and streaming unit for £1.6 billion
  • June 29, 2026: Comcast announces the separation of NBCUniversal and Sky
  • July 2026: Liberty Sky Advisors publishes A Risk Too Far, setting out the recoup asymmetry and the 15p figure

Summary

Who. Spend-tracking firms calculate it, with Guideline the source of the published television figures through its agency and large-independent client base. Broadcasters are the subject, agency holding companies control the budget lines it depends on, and Google, Meta and Amazon are where the unrecouped share lands.

What. The share of advertising money leaving one medium that returns to the same sellers elsewhere. For UK television, roughly 15p in the pound returns on broadcaster streaming, against about twenty-five cents in the dollar in the United States and twenty cents in Canada.

When. The ratio entered public debate in July 2026 through the Liberty Sky Advisors assessment A Risk Too Far, drawing on Guideline data discussed earlier that year on the Media Unfiltered podcast.

Where. In the trading structures that separate linear, broadcaster streaming, digital video and connected television into distinct budget lines, and in the United Kingdom policy consultation on ending terrestrial broadcast in 2034 or 2044.

Why. Broadcasters defend their economics on the claim that streaming recaptures what linear loses. The recoup rate is the arithmetic test of that claim, and at roughly five times the linear price for broadcaster streaming against a YouTube rate sitting inside the linear band, the current answer is that most of the money does not come back.