The largest structural change to the US connected television advertising market in years arrived on June 15, 2026. Fox Corporation and Roku announced a definitive acquisition agreement valuing Roku at approximately $22 billion in enterprise value, with Fox paying $160 per Roku share in a combination of $96 in cash and 0.9693 shares of Fox Class A common stock. The deal, reported on the same day by PPC Land, AdExchanger, Digiday, Adweek, and MediaPost, represents a 34 percent premium to Roku's unaffected share price as of June 11 and a 21 percent premium to its unaffected 52-week high. Shareholders in both companies must approve it, as must US regulators under the Hart-Scott-Rodino Act. The transaction is expected to close in the first half of 2027.

The deal inverts a six-year chapter of separation. Fox sold its stake in Roku to fund the March 2020 acquisition of Tubi for $440 million. At the time Tubi had a share of ad-supported streaming that was roughly comparable to The Roku Channel. By Q4 2025, per Nielsen Gauge data, The Roku Channel held a 6.3 percent share of all US ad-supported streaming while Tubi sat at 6.2 percent. Together, on a pro forma basis, those two platforms plus Fox's live sports, Fox News, and entertainment rights would constitute the third-largest player in US television by share of viewing, behind only YouTube and Netflix.

That is the strategic frame Fox presented to investors on June 15. Lachlan Murdoch, executive chair and chief executive of Fox Corporation, called the news a "defining moment," arguing on an analyst call that combining Roku's first-party data from more than 100 million global streaming households with Fox's live sports portfolio and Tubi's AVOD catalog creates an advertising offer no pure-play broadcaster can match. "This combination will transform the scope of our company," Murdoch said. Roku's founder and chief executive Anthony Wood is set to join Fox's board after closing.

The financial architecture is notable. Roku disclosed its advertising revenue separately for the first time in Q1 2026: $612.7 million for the quarter, up 27 percent year over year. Subscription revenue grew 30 percent to $518.5 million in the same period, bringing total Q1 net revenue to $1.25 billion. Full-year 2025 advertising revenue was $2.4 billion, with total revenue guidance for 2026 set at $5 billion before the deal was announced. Madison and Wall, cited by MediaPost, estimated that the combined company will generate around $9 billion in total advertising revenue, representing roughly 14 percent of all US advertising spend and approximately 16 percent of US streaming ad spending.

That figure changes the competitive arithmetic of the programmatic CTV market dramatically. AdExchanger noted that Fox would own both the number one (The Roku Channel) and number two (Tubi) most popular free streaming services in the United States simultaneously. Digiday reported that industry analyst Brian Wieser told reporters the deal gives Fox a formidable blend of reach, first-party data, and live programming that makes the combined entity far more attractive to advertisers than either business separately. The Roku platform held a 32 percent share of open programmatic CTV inventory in February 2026, with Amazon Fire TV at 16 percent, Apple TV at 15 percent, and Samsung at 14 percent.

Both companies committed in their SEC filings to operating Roku as an open, partner-friendly platform. Competing streaming services currently distribute through Roku's operating system and depend on its homescreen for discovery. Whether that neutrality will be maintained after a competitor owns the OS is the question Adweek put directly to ad industry experts on June 15, reporting that three major questions emerged: how data sharing will work across Fox and Roku's combined first-party identity graphs; whether Roku's existing DSP partnerships and Amazon data deal will survive under Fox's ownership; and whether the deal will trigger further consolidation from rivals who will now face a competitor with far greater scale and vertical reach.

The answer to that last question may already be emerging in the data. PPC Land reported on June 14 that political CTV spending projections for the 2026 midterm cycle had risen to $2.7 billion, per AdImpact estimates, up from a previous estimate of $2.4 billion. In that environment, Fox-Roku's scale in both live news and streaming creates a concentrated sell-side position for political advertisers, who prize authenticated household data and premium brand environments.

The Roku platform architecture that Fox is acquiring has several technically distinct components that are worth separating. The operating system layer is installed on devices from dozens of hardware manufacturers beyond Roku's own branded hardware, giving it broad distribution across television sets it does not sell directly. The homescreen, which is where Roku sells high-CPM display and video inventory, sits above the OS and is a direct-sold asset that has not previously been available programmatically at scale. That is changing: PPC Land reported on June 10 that Samsung Ads was opening its own smart TV homescreen to programmatic buyers via The Trade Desk and Google DV360, powered by Magnite's SpringServe, rolling out globally in Q3 2026. Fox acquiring Roku's homescreen before the rest of the industry has established a programmatic channel for smart TV homescreens gives it a defensive position that is difficult for a purely programmatic buyer to replicate.

Roku's OneView demand-side platform, which allows advertisers to buy across The Roku Channel, Tubi once integrated, and third-party streaming services distributed through the Roku OS, runs on a first-party identity graph that is authenticated at account level across all 100 million-plus streaming households. That identity layer is the asset Lachlan Murdoch highlighted on the analyst call. Most CTV measurement today depends on probabilistic household matching or panel-based extrapolation. Roku's authenticated graph is deterministic. The combination with Fox's content creates a circumstance where one company controls the content, the platform, the distribution, and the identity layer simultaneously across a footprint that covers roughly half of all US broadband households.

From a programmatic infrastructure standpoint, PPC Land had reported on June 13 that Roku was exploring a sale attracting media and technology buyers at a valuation around $19 billion. The $22 billion final figure exceeded those early reports, and the 20 percent stock surge Roku saw on June 12 on initial deal rumours had already reset expectations before Monday's official announcement.

Publicis and The Trade Desk settle, and say almost nothing about it

The second major story of the period arrived three days earlier with less fanfare but nearly as much structural significance. Publicis Groupe and The Trade Desk issued a joint statement on June 12, 2026, announcing that they had resolved their public dispute. Publicis reinstated The Trade Desk as a recommended DSP for clients. The statement was short. Neither party disclosed the terms, the concessions, or which side moved.

Digiday covered the resolution on June 12, characterising it as either a triumph of negotiation or a sign that the confrontation was never quite as dramatic as it appeared. PPC Land covered the broader context in its June 13 newsletter, and MediaPost reported the resolution on June 12, noting that The Trade Desk's stock had dropped around 13 percent when Publicis pulled it from the recommended DSP list in March.

What started the dispute matters for understanding the settlement's silence. In March 2026, Publicis commissioned an audit through FirmDecisions that found, in Publicis's characterisation, irregularities in how The Trade Desk applied its fees. Specifically, Publicis alleged The Trade Desk was stacking its own ad tech fee on top of other charges in a way that Publicis said was not supported by their contract. Publicis told clients to stop spending with The Trade Desk's platform. The Trade Desk denied all allegations. For a company whose entire value proposition is transparency in programmatic media buying, the public framing of the dispute was particularly damaging.

The resolution changes the competitive picture slightly but leaves the underlying tension unresolved. Digiday reported on June 15 that DSP competition is intensifying across the board, with The Trade Desk, the sector's largest independent platform, under sustained pressure from Amazon, while Yahoo, StackAdapt, and Viant are all attempting to gain share. Buyers Digiday spoke to on condition of anonymity described CPM rates, access to log-level data, and streaming partnerships as their primary evaluation criteria. One said the preference was to stay "as agnostic as possible," which implies that the Publicis-TTD feud may have had less operational impact on buying behaviour at the desk level than the public statements suggested.

The resolution also arrives at a moment when The Trade Desk's TAG fraud certification had lapsed, as PPC Land reported in its June 14 newsletter. Google's TAG certification had lapsed around the same time. Both companies' credentials in the Trustworthy Accountability Group's registry were showing as expired. The lapse of fraud certification does not immediately affect campaign delivery, but it undermines the signal of trust that TAG accreditation is designed to provide. For a platform whose brand is built on supply chain transparency, renewing that certification will be a priority.

What the DSP scorecard published by Digiday on June 15 reveals about the competitive landscape is relevant here. Buyers describe The Trade Desk as the default choice for open programmatic CTV but increasingly question whether it maintains its premium as Amazon's DSP closes the gap on inventory breadth and as smaller platforms like StackAdapt demonstrate faster product velocity on newer ad formats. The Publicis dispute, whatever its private resolution, briefly opened a gap that competitors attempted to enter. The joint statement closes that gap publicly, but the underlying structural question, whether a single independent DSP can sustain the market position The Trade Desk has built as the walled gardens build their own full-stack alternatives, remains unanswered.

The settlement's silence on terms is itself informative. Both parties have an interest in not publicising which side conceded. Publicis cannot afford to be seen to have launched a public campaign with financial consequences for The Trade Desk's stock price and then retreated quietly. The Trade Desk cannot afford to admit to the fee irregularities Publicis alleged without creating liability exposure from other holdcos that use its platform. The joint statement's brevity is therefore not a failure of communication; it is the only statement both parties could issue without creating additional damage to one or both of themselves.

UK publishers build a billing mechanism for AI scrapers

On June 15, 2026, UK publishers launched Search-Only Contracts, a legal mechanism designed to charge AI companies £500 per article scraped. PPC Land reported the initiative on June 15, describing how the mechanism uses county courts rather than intellectual property litigation. The choice of venue is deliberate. IP litigation requires specialist counsel, years, and significant capital. County court small claims are accessible to individual publishers without legal representation, capped at amounts that make individual cases viable.

The contractual logic is that when an AI crawler accesses a page under terms explicitly permitting access only for the purpose of being indexed in search results, any use beyond that, including training a large language model or synthesising an answer to a user query, constitutes a breach of those terms. The £500 figure is a standing per-article tariff that makes the economic case for enforcement straightforward: a publisher with 10,000 scraped articles could theoretically pursue a £5 million county court claim.

The initiative connects to a legislative front that had been advancing separately. New York's Assembly passed bill A11292 on June 5, 2026, requiring AI crawlers to disclose their identity and purpose to news publishers or face $15,000 per day penalties for non-compliance. Search Engine Roundtable cited the Press Gazette on June 15 as noting that publishers were preparing to take AI companies to court if they did not pay under similar contractual arrangements. The New York bill and the UK contractual mechanism address the same gap from different angles: the New York bill forces disclosure so publishers can identify which crawlers they are dealing with; the UK contracts provide a remedy once a scraping event has been documented.

Why this moment? The traffic loss is now severe enough to have closed sites. PPC Land reported on June 14 that All About Berlin, an eight-year-old independent expat guide, had lost 70 percent of its organic search traffic to Google AI Overviews. On the same day, PPC Land also reported that Overfishing.org, a 21-year-old independent conservation site, had shut down entirely after its traffic was absorbed by Google AI Overviews. These are not traffic fluctuations. They are site closures that follow directly from AI systems consuming and synthesising content whose publishers can no longer sustain their operations.

The enforcement challenge for county court claims is documentation. Publishers must be able to show that a specific crawler accessed a specific page on a specific date and that the terms of access prohibited the use made of the content. Most publishers do not collect web server logs with the granularity required to attribute a scraping event to a particular AI crawler with certainty. That documentation gap is addressable through technical means: publishers can deploy canary content, generate unique tracking tokens in specific page URLs, or use honeypot pages that are only discoverable by automated crawlers and not by human users. The UK initiative is therefore likely to be most effective for technically sophisticated publishers who have proactively instrumented their server logs, which is a smaller population than the total set of publishers with a legitimate grievance.

The £500 per-article figure also invites attention to scale. OpenAI and Google process billions of web pages in their training pipelines. At £500 per article, a publisher with 5,000 scraped articles could claim £2.5 million in a county court proceeding, a sum that exceeds the county court's small claims limit of £10,000 and would require the multi-track procedure with its associated costs. The scheme as designed works well for smaller publishers with smaller article counts. For the largest claims, the economics push claimants toward specialist IP litigation, which is precisely the route most publishers have tried to avoid. The mechanism may therefore be most effective as a deterrent and as a precedent-builder in small cases, rather than as a comprehensive revenue recovery system for the industry at scale.

Search Engine Roundtable linked on June 15 to a Glenn Gabe analysis on X showing that publishers who had already lost rankings in Google Search subsequently lost visibility in AI Overviews and ChatGPT as well. The dependency compounds: lose organic visibility, lose AI citation visibility. Lose both, and revenue collapses without a meaningful path back. Search Engine Roundtable also cited Similarweb data showing that ChatGPT advertiser count had grown 46 percent in a single week in mid-June, driven by a lower spend threshold for participation, which underscores that OpenAI is building an ad-funded model on top of content its systems helped defund.

Perplexity's position in this landscape is particularly pointed. PPC Land reported on June 15 that Perplexity co-authored a study claiming AI agents cut task time by 87 percent, while copyright and security cases mounted against the company. The study used Perplexity's own data and products to reach its conclusions, a self-referential structure that drew scepticism from researchers examining the methodology. Publishers pursuing the UK Search-Only Contracts will be unlikely to be reassured by vendors producing their own validating evidence under those circumstances.

Three updates from Google Ads were announced on June 15 and converge on August 17, 2026 as a key change date. PPC Land reported all three. Search Engine Roundtable's June 15 daily recap flagged the same cluster, with Barry Schwartz noting that the bidding target change generated the most discussion among practitioners.

The first change is the expansion of Smart Bidding Exploration to Performance Max campaigns. Smart Bidding Exploration allows the bidding algorithm to test outside current targets in search of additional conversion volume the existing strategy would not capture. Performance Max already operates with significant automation: campaigns span search, display, YouTube, Discover, Gmail, and Maps simultaneously, with asset groups rather than ad groups, and placement reporting is deliberately limited. Layering exploration behaviour on top of an already opaque campaign type means advertisers will see their budgets probing further outside their target parameters with less visibility into where the exploration is occurring.

The second is a promotion mode beta, initially US-only, that gives advertisers a dedicated interface for setting promotional rules on campaigns. Instead of encoding a sale period through temporary bid adjustments or target changes, advertisers can configure promotion-specific logic that activates and deactivates on a schedule without disturbing the base bidding strategy. That separation is operationally useful for retailers who manage recurring promotional cycles.

The third is the bidding target optimisation change, which Google describes as delivering more predictable performance for campaigns with fixed spending limits. Search Engine Roundtable's reporting on June 15 raised the concern that "fixed spending limits" constrains the budget but does not constrain how Google allocates bids within that budget across placements. An advertiser running a target CPA campaign may find the August update treats that target as a directional goal rather than a ceiling. Google has not published implementation specifics beyond the policy announcement.

The three August changes sit alongside two additional developments the Search Engine Roundtable recap documented for June 15. Google is rolling out Information Agents within Google Search AI Mode to paid Google AI Ultra subscribers. Information Agents, announced at Google I/O 2026, run background searches on behalf of users even when they are not actively querying, returning results when relevant signals emerge. The feature is subscription-gated, but its existence means Google's paid search index is now being actively queried by background agents on behalf of opted-in users, which changes the interpretation of query volume data for advertisers. An impression served against a query initiated by an agent on behalf of a human subscriber is a qualitatively different interaction than one triggered by a human sitting at a keyboard. Google has not yet published guidance on how agent-initiated queries will be attributed or counted in the Google Ads auction.

Separately, Google sent email notices to some advertisers on June 15 about the expansion of its limited ad serving policy to Google Search. PPC Land had reported this development on June 13. The policy, previously applied to other Google properties, limits ad impressions from what Google terms unqualified advertisers on queries it believes are associated with negative ad experiences. The enforcement is graduated rather than binary and phases in from June 2026 through 2028. The practical effect is that new or low-quality advertisers will see impression throttling on sensitive query categories before the policy's full enforcement window closes.

The full picture of Google's search advertising platform in mid-June 2026 also includes Google Shopping testing a "Good Price" label in product carousels, as reported by Search Engine Roundtable on June 15, and a bug in Google Business Profile where invitations to add new owners and managers were failing to deliver. Search Engine Roundtable flagged the Business Profile access bug on June 15 as significant for businesses that manage local listings across multiple administrators. Google had also announced the integration of Gemini with Google Business Profile, as PPC Land reported on June 13, adding Business Notebooks that surface proactive alerts for local businesses without requiring manual query input.

The DSA-to-AI Max migration delay, reported by PPC Land on June 13 and confirmed by Search Engine Roundtable on June 12, pushed the automigration deadline from September 2026 to February 2027. The delay restores DSA campaign creation, which Google had disabled ahead of the original migration date. Advertisers who had been preparing for September now have a longer runway, but the February 2027 deadline is firm. The combination of the DSA delay, the August bidding changes, the Information Agents rollout, and the limited ad serving policy expansion represents a structural reshaping of how Google Search monetises across multiple surfaces simultaneously, compressed into a window of roughly eight weeks.

Streaming wars: VAB data, agentic AI in healthcare, and Uber's retail media push

Against the Fox-Roku backdrop, several other stories on June 15 mapped the conditions that make the deal's advertising projections plausible.

VAB published June 2026 streaming research showing that 89 percent of US streamers now use ad-supported services. FAST channels had reached 1,700 titles across available platforms. Interactive CTV ad formats were producing 138 percent higher brand recall compared with standard pre-roll units. VAB's addressable TV guide for 2026, published on June 14, found that 92 percent of US pay TV households were addressable-enabled, with nearly half of current addressable advertisers raising budgets this year. Fox-Roku will enter this market as the largest single addressable household footprint outside the walled gardens.

DeepIntent launched Helix AI on June 15, described as the first agentic platform for healthcare marketers, claiming to halve media planning cycles through HIPAA-compliant natural language tools. The product arrives as pharmaceutical brands are among the heaviest CTV buyers: PPC Land had reported on June 13 that DeepIntent had already launched verified live CTV inventory, custom pacing, and pharma KPI measurement for healthcare brands targeting World Cup, NFL, and NBA audiences. Helix AI automates the planning step that precedes that buying activity, using natural language prompts to generate audience strategies and media schedules rather than requiring planners to build them manually in a media planning tool.

Uber Advertising launched three new products on June 15: Offers on Uber, which places sponsored incentives in the rideshare booking interface; new premium formats on Uber Eats including brand takeover placements; and Offsite Ads, which extend Uber's first-party audience data to third-party inventory placements for the first time. Offsite Ads is the most structurally significant. It creates an audience export mechanism that allows brands to reach Uber users on inventory outside the Uber platform, using Uber's first-party data on mobility patterns, food ordering behaviour, and location history. That is the retail media model transposed to a mobility platform.

The retail media measurement problem these products inherit was quantified on the same day. PPC Land reported a meta-study by Incremental covering 150,000 campaigns and $350 million in ad spend, finding that siloed attribution in retail media systematically misses between 36 percent and 53 percent of true campaign ROI. Uber's Offsite Ads will produce conversions that register in third-party attribution systems rather than Uber's, unless advertisers construct dedicated cross-channel measurement architectures. The measurement problem does not diminish the strategic value of the product, but it will complicate how buyers evaluate and justify the incremental spend.

Microsoft Advertising's June 15 announcement of job seniority targeting via LinkedIn profile data, across 10 seniority levels in 29 markets including the US and EMEA, addresses a distinct segment of this audience-data conversation. The LinkedIn targeting expansion gives B2B search advertisers a way to filter impressions by the organisational rank of the person conducting the search, not just their job function or industry. The practical effect is that a query like "enterprise procurement software" can now carry bid differentials depending on whether the person searching is at director level or above. That is a meaningful shift for B2B advertisers whose conversion rates are highly sensitive to seniority.

Zattoo's June 15 announcement that it had named Stro-er as its exclusive ad sales partner for German CTV inventorycovering smart TVs, streaming players, and mobile devices illustrates the European parallel to Fox-Roku. Where US consolidation is proceeding through billion-dollar acquisitions, European markets are consolidating through commercial outsourcing arrangements. Smaller streaming platforms that cannot sustain internal ad sales teams are delegating monetisation to established media sales houses with existing agency relationships. The model serves the same purpose at a different scale.

Publisher trust, ad fraud, and the DSP market under pressure

The ad tech trust layer showed further stress in the same 48-hour window. PPC Land's June 14 newsletter documented that TAG's fraud-prevention certifications for both Google and The Trade Desk had lapsed, at the same time that Lunio's Q1 2026 analysis of 64 million ad clicks had found LinkedIn's invalid traffic rate at 17.62 percent. Bing's IVT rate was 12 percent. Google Display Network IVT had surged 132 percent year over year. One placement Lunio identified had an IVT rate above 93 percent. PPC Land covered the Lunio analysis separately on June 13.

The Digiday DSP scorecard published on June 15 found that buyers evaluating platforms weight transparency of log-level data, CPM rates, and CTV inventory quality as their primary criteria. Against that background, The Trade Desk's expired TAG certification and Publicis's allegations about fee stacking both land as transparency failures in the criteria that most matter to buyers. The Publicis-TTD resolution returned the relationship to normal without resolving those transparency questions publicly.

PPC Land's June 14 newsletter on agentic AI traffic added a related data point. HUMAN Security's May 2026 data showed agentic traffic across the web fell 4.3 percent month over month, while publisher blocking rates rose to nearly 9 percent. As more publishers actively block AI crawlers, the data scarcity problem for AI training corpora grows. That dynamic pushes AI companies toward licensing deals with publishers who choose to participate, which is the other half of the publisher economics conversation that Search-Only Contracts and the New York disclosure bill represent.

The Brainlabs perspective on the agency landscape, published by PPC Land on June 15, argued that independent agencies are capturing high-margin, AI-driven work that legacy holding companies are ceding because their internal structures are too slow to adapt. That argument, from Brainlabs Chief Growth Officer Ali Reed, positions the Publicis-TTD dispute as a symptom of a larger transition: the holding company model is under structural pressure from platforms that can deliver performance automation directly, and the disputes about fees and transparency are partly driven by agencies trying to preserve margin in a stack that is increasingly bypassing them.

HubSpot's partner revenue projections on June 15 offered a data point from a different angle. IDC projected $42 billion in HubSpot partner revenue by 2030, with AI-first revenue growing at a 28.4 percent compound annual growth rate. A separate HubSpot survey of more than 3,000 CRM buyers found AI search had become the single strongest predictor of purchase intent, ahead of demos, peer reviews, and sales calls. PPC Land reported that HubSpot data on June 13. The implication for both advertisers and agencies is that AI-mediated discovery is becoming the primary entry point in B2B buying journeys, which changes both the channels that matter and the role of human intermediaries.

Also noted