There are days in the ad tech calendar where the volume of significant news compresses a month's worth of structural change into a single cycle. June 10, 2026 was one of those days. A Munich civil court issued what may be the most consequential judicial ruling in the short history of AI-generated search results. Samsung opened a category of premium TV inventory to programmatic buying that has never been available that way before. New data from AppsFlyer redrew the map of where mobile ad fraud now concentrates. ChatGPT's advertising CPMs were documented falling sharply from their launch levels. A new study confirmed that 68 percent of Google searches now end without a click. And across more than a dozen other announcements, the industry's measurement infrastructure, agentic standards, audience data pipes, and streaming distribution landscape all moved simultaneously.

The Munich ruling: Google is directly liable for what AI Overviews generates

On May 28, 2026, the Landgericht Munchen I, Munich's regional civil court, issued a preliminary injunction under case reference 26 O 869/26 holding Google directly and primarily liable for false and defamatory claims generated by its AI Overviews feature. The court's 26th civil chamber, following an oral hearing on April 23, found that Google bears authorial responsibility for the accuracy of text the system produces. Violations carry fines of up to EUR 250,000 per instance or custodial sanctions. Google must bear 80 percent of court and legal costs.

PPC Land reported the ruling on June 10, drawing on the full court documentation. The case was brought by two German publishing companies forming part of a Munich-based media group operating 12 branded imprints. The second plaintiff is a subsidiary publishing books and magazines under the GeraMond brand, covering technology and history. The triggering queries combined the companies' names with the word "Betrugsmasche," German for fraud scheme. AI Overviews generated false and defamatory text in response. The harmful responses were first documented on January 20, 2026. Formal notice went to Google on February 2. The hearing took place April 23. Judgment arrived May 28, nearly four months after the publishers first identified the problem.

The legal architecture of the ruling is its most significant element. Google has consistently argued in every prior case that it functions as an intermediary, surfacing and distributing information rather than authoring it. That argument has worked in search engine liability cases for two decades because search engines indexed and linked to content created by others. AI Overviews does not index. It generates new text. The Munich court applied that distinction directly: when a system produces new text, the operator of that system bears authorial responsibility for the accuracy of what it produces. Hosting a link to a false claim is not the same as generating the false claim. The court found generation is not intermediation.

This finding converts a conditional possibility into a binding judgment. A Frankfurt Regional Court ruled in September 2025 that AI Overviews can materially harm competition and reduce click-through rates, and stated that an objectively incorrect AI Overview "could constitute an unreasonable obstruction" under German competition law. That conditional finding, applied to a case where the summary was found "ultimately not false," produced no remedy. Munich took the same legal reasoning and applied it to a case where the falsity was documented and undisputed.

The broader litigation context matters enormously here. In July 2025, the Independent Publishers Alliance filed a formal antitrust complaint with the European Commission alleging Google misuses publisher content to build AI Overviews without compensation or an opt-out mechanism. The European Commission opened separate Article 102 proceedings in December 2025 on the same grounds. In the United States, Penske Media Corporation, owner of Rolling Stone, Variety, and Billboard, filed a federal antitrust case in September 2025. A Minnesota solar company filed a separate US suit over specific false AI-generated claims about its business. The Helena World Chronicle case, dismissed by a US federal judge in April 2026, turned on antitrust theory rather than content liability, leaving the direct defamation route in American courts untested.

The traffic damage underlying all this litigation is quantified. Ahrefs published data in February 2026 showing AI Overviews correlate with a 58 percent reduction in click-through rates for top-ranked pages. Earlier data from March 2025 measured a 34.5 percent organic click reduction for pages in results containing an AI Overview. Organic search traffic to news publishers fell from 51 percent to 27 percent of their total traffic between 2023 and 2025.

That last number connects directly to a Search Engine Roundtable report from June 10. Rand Fishkin and collaborators at SparkToro and Similarweb published an updated zero-click study showing that 68.01 percent of Google searches ended without a click in the most recent measurement period. When filtered specifically for queries that received an AI-generated response, the share of traffic going to the open web dropped to just 27.6 percent. Barry Schwartz reported these findings on Search Engine Roundtable on June 10. The data are not from a point-in-time snapshot. They are part of an ongoing study, and the trend is in one direction. The Munich court has now established, at least at the preliminary injunction level, that the platform generating those summaries bears direct responsibility for their accuracy. Any search practitioner, publisher, or brand legal team watching these numbers in parallel now has a precedent to work with.

The ruling is a preliminary injunction, not a final judgment. It is limited to specific false claims about specific plaintiffs. Google has every right and resource to contest it through the full appeals process. But it establishes that a European civil court will hold the operator of an AI-generated feature directly responsible for the text that feature produces, on financial penalty terms. That is a materially new operating condition for any platform generating summaries at scale, and it arrives at a moment when the most important measurement story in search is how little of the traffic generated by those summaries ever reaches the sites whose content informed them.

Samsung opens its home screen to programmatic buyers

Samsung Ads announced on June 10 that its Smart TV home screen placements are now available through programmatic buying for the first time. The Trade Desk and Google DV360 are the two DSP integrations at launch. Magnite's SpringServe ad server powers the supply side. Global rollout begins in Q3 2026, with additional DSP partners expected to follow.

Home screen placements are structurally different from every other CTV inventory category. They appear before any app opens, before any content plays, at the moment a viewer switches on the set. Audience attention at that moment is undivided and pre-committed: the viewer has not yet chosen what to watch, has not navigated into a stream, has not been served a pre-roll. Until June 10, reaching that surface required a direct deal with Samsung Ads. The announcement places it inside the same workflow buyers use for the rest of their programmatic CTV campaigns.

AdExchanger reported on June 10 that one of the central challenges in bringing this inventory to programmatic buyers has been packaging it: how do buyers evaluate reach and frequency for a unit that does not fit the format categories that standard measurement infrastructure was built around? Samsung and The Trade Desk are actively working with third-party measurement companies to address this. The IAB is developing CTV ad portfolio standards, and Samsung said it did not want to wait for that process to complete. Magnite Group SVP Mike Laband described SpringServe's role as making "high-impact CTV placements more seamless, scalable, and measurable."

The Magnite relationship did not start on June 10. Samsung and Magnite expanded a partnership around audience addressability in streaming television in April 2025. Magnite merged its SpringServe ad server with its SSP infrastructure in April 2025, creating a unified platform that Jounce Media's March 2025 Supply Path Benchmarking Report credited with 99 percent coverage of connected television supply in the United States. Inside DV360 reporting, this combined entity now appears as Magnite SpringServe, following a November 2025 naming update. The Samsung home screen announcement is the latest extension of an infrastructure that has been consolidating around the SpringServe stack across an increasingly wide range of publishers and platforms.

Samsung is not the first TV OS provider to move home screen inventory toward programmatic. LG also presented its home screen as premium ad inventory at NewFronts in March 2026. The distinction that matters commercially is scale: Samsung's global installed base is one of the largest in the world, and the home screen is the highest-attention surface on the device.

Google is overhauling DV360 with 12 changes by August

A DV360 platform roadmap last updated May 22, 2026 details more than 12 changes arriving between now and August, spanning audience targeting, brand safety, deal creation, reporting infrastructure, and creative rotation. Several have already taken effect. The roadmap is the densest public change schedule the platform has published in a single quarter in recent years.

The June wave is immediately operational for active campaigns. Legacy audience expansion for YouTube campaigns is deprecated around June 15, replaced by lookalike audiences that require manual setup. There is no automatic migration of existing configurations. Teams that have relied on audience expansion as a background optimisation layer for YouTube Video Reach or Video View campaigns and have not yet built lookalike audience configurations will find that targeting mechanism absent from the date of deprecation. The Data Onboarder dimension in DV360 reporting renames from "AdAdvisor by Neustar" to "AdAdvisor by TransUnion," following a corporate rebrand at the data provider level. Any saved reports, dashboards, or automated workflows that reference the Neustar label will need to be updated. Floodlight and Inventory Availability reports begin a phased migration from legacy offline reporting to instant reporting, creating a transition window where data behaviour may differ depending on timing of access. The phased nature of the migration means that two users pulling the same report on different days within the transition period may receive data from different systems.

The August wave is structurally heavier and requires more preparation time. Digital content labels, the brand safety exclusion taxonomy in continuous use inside DV360 and Campaign Manager 360 since the platform's early years, are deprecated. Sensitive category exclusions for display, video, CTV, and audio line items go with them simultaneously. The replacements are inventory Modes and content themes. There is no automatic mapping from old label structures to new ones. Every brand safety exclusion configuration built around the content label taxonomy needs to be rebuilt in the content themes architecture before August. Advertiser-level category exclusions including News and Politics also retire in August, replaced by content themes, with the same absence of automatic migration.

The deal creation workflow changed in April: manual non-programmatic guaranteed deal creation was deprecated, with the Deal Sync API now the sole creation method for new deals. PPC Land has tracked this direction across multiple prior DV360 updates, including the May 2026 launch of Partial Structured Data File Upload, which let advertisers update line items with far fewer than the standard 120-plus columns required. The pattern is consistent: from manual, UI-based controls toward API-driven, signal-specific mechanisms.

One separate but directly related change completed its rollout on June 10 specifically: the DV360 API finished its phased deployment of full Demand Gen support for all partners. The rollout started June 10 and was scheduled to reach every DV360 partner by June 24. It closes a gap that had required manual UI work for certain Demand Gen campaign operations since the format was introduced, and is particularly relevant following Google's expansion of Demand Gen product feeds to the automotive vertical at Google Marketing Live 2026. All four Demand Gen ad format types, covering video, image, carousel, and product ads, are now fully scriptable within the API.

AppsFlyer: organic traffic is now the largest single source of mobile install fraud

AppsFlyer's State of Fraud for Marketers 2026, published June 10, covers 106.4 billion total installs across 246,000 apps from Q1 2025 through Q1 2026. Organic traffic now accounts for 52 percent of all fraudulent mobile installs, more than any paid channel combined.

That headline figure is significant not as a standalone data point but because of what it does to the rest of mobile measurement. Organic install rates serve as the internal baseline against which every paid channel is evaluated. If a paid campaign appears to perform well, the comparison point is what organic delivers without paid intervention. Inflate the organic number, whether through deliberate fake installs constructed to look like self-discovered traffic or through attribution failures that push paid installs into the organic bucket, and every paid channel's measured performance becomes systematically overstated. The fraud is not simply wasted spend. It corrupts the measurement baseline against which every other number is read.

DSP fraud grew 59 percent year on year. Owned media fraud, covering email and push, grew 221 percent. Spoofing was the fastest-rising fraud technique across 2025, outpacing total install growth every quarter between Q3 2025 and Q1 2026. Unlike install hijacking, which steals attribution credit for genuine installs that were already occurring, spoofing fabricates everything from scratch: devices, users, behavioral event sequences, and timing patterns designed to pass standard verification checks. AppsFlyer describes it as the type of fraud most likely to be undercounted in any dataset, including the one underpinning the 2026 report itself, precisely because it generates clean-looking signals rather than anomalies.

The vertical data in the report is granular and, in some categories, severe. In Finance on Android, Real Users Lift, the metric measuring what share of reported installs reflect genuine human users, has been locked between 50 and 53 percent for five consecutive quarters without improvement. Half of what Finance advertisers believe they are acquiring on Android does not exist, and that has not changed since Q1 2025. In Social Media on iOS, Real Users Lift reached 275 percent in Q2 2025, meaning three in four installs were fraudulent for an entire quarter. Every ROAS figure, cohort analysis, growth metric, and retention curve built on Social Media iOS data uring Q2 2025 was measuring an audience that was largely absent. The affiliate-to-self-reporting-network fraud gap stands at 36 times, concentrating risk squarely on media plans that use affiliate channels with limited third-party oversight.

ChatGPT CPMs are falling, and LiveRamp just built the measurement infrastructure to prove it

Digiday reported on June 10 that ChatGPT advertising CPMs have fallen sharply from their launch levels. The rate advertisers pay to reach every thousand users dropped from $60 at the platform's launch ten weeks ago to as low as $25 in some cases. The trend is downward. When impression prices fall, so does the revenue each impression generates. This data point matters because the $60 CPM was the figure OpenAI used to position ChatGPT ad inventory as comparable to premium streaming and NFL broadcast placement. The gap between that positioning and the actual market clearing rate is now measurable.

On the same day, LiveRamp announced that its Conversions API Hub now connects to OpenAI's ChatGPT, giving advertisers running ChatGPT campaigns a server-to-server path for conversion signal delivery. The announcement was published by Travis Clinger, Chief Connectivity and Ecosystem Officer at LiveRamp. The CAPI architecture routes conversion event data from an advertiser's own server infrastructure directly to OpenAI's systems, bypassing browser-based tracking entirely. ChatGPT's conversational interface creates fragmented signal conditions: users move between sessions, devices, and interaction modes in ways that make client-side tracking structurally unreliable. Server-to-server connections remove that dependency. Meta's own data shows CAPI users see a 17.8 percent lower cost per result on average compared to those relying solely on client-side tracking.

The LiveRamp integration arrives as ChatGPT's ad platform is building out rapidly. Conversion-optimized campaigns launched June 5, with eligibility gated on having an active conversion event flowing through the pixel or CAPI by June 1. Daily budget controls and ZIP-code-level geographic targeting arrived May 22. Self-serve Ads Manager opened to all US businesses May 5 with CPC bidding at a recommended $3 to $5 per click. The platform launched its pilot February 9 with a minimum spend requirement of $200,000 to $250,000. That floor was eliminated entirely. The progression from a locked premium pilot to a self-serve performance platform with server-side measurement infrastructure took four months.

Digiday also noted on June 10 that OpenAI has already burned through $2.5 billion in 2025 and projects that figure rising to $8.5 billion this year. Ad revenue of $102 billion has been modelled by analysts as a possible long-run ceiling, though the path from current CPM compression to that number is long and unresolved. The tension between pricing erosion at the impression level and the cost base underlying the platform is the central commercial question OpenAI's ad business now has to navigate.

The upfront is moving toward outcomes, agentic agencies, and the infrastructure that runs them

Digiday's Future of TV Briefing on June 10 examined how this year's upfront is setting up outcome-based buying as the structural model for future negotiations. TV and streaming ad sellers have moved toward guaranteeing business outcomes rather than just reach and frequency metrics, and this year's negotiations are the first in which that shift has moved from a negotiating position to a contractual norm across multiple major sellers simultaneously. The specific buyers and sellers who have moved furthest, and the terms of those guarantees, remain behind Digiday's member paywall, but the directional finding matters in its own right: the upfront is being rewritten around accountability to results, not delivery of impressions, and that creates a different measurement conversation between buyers and sellers than the one that has governed the market for decades.

A separate Digiday June 10 piece on agency evolution documented that brands and their agency partners have now begun using AI agents to plan, execute, and optimise ad purchases in production environments, not just pilots. The agentic advertising transition documented across Digiday, AdExchanger, and PPC Land throughout 2026 has been theoretical for most of the year. The June 10 piece is one of the clearer signals that capital and workflow have aligned behind actual deployment. The Trade Desk launched its in-platform Koa Agents in April 2026 with Stagwell as pilot partner. Adobe introduced AI agents for digital marketing the same month. Amazon has run an Ads Agent for natural-language campaign management since November 2025. The competitive pressure to offer agentic capabilities is no longer prospective.

That pressure connects directly to Affinity joining AdCP as a Founding Member on June 10, pushing the Ad Commerce Protocol's standard-setting work to cover OEM, browser, app store, and TV OS surfaces that programmatic has never reached. The Samsung home screen announcement and the AdCP expansion both represent inventory surfaces that existed outside the programmatic stack being pulled into it, through different mechanisms but toward the same end. Samsung is doing it through DSP integrations with established technology. AdCP is building the governance layer for surfaces where there are no established DSP integrations yet.

Digiday's LinkedIn piece on June 10 reported that the platform is working to make B2B creator discovery more scalable for brands looking to activate creator content on the professional network. LinkedIn's creator ecosystem has grown significantly in 2025 and 2026, and the platform is building tooling to make that inventory findable and activatable at speed comparable to what brands expect on consumer social platforms. This sits alongside the broader creator economy infrastructure story of 2026: platforms competing to be the discovery layer between brands and creators, with data and programmatic tooling replacing personal relationships as the primary activation mechanism.

Google's own search data documents the NBA Finals search surge

Google Trends data published on June 10 shows the Knicks versus Spurs NBA Finals matchup as the top US search query, with vintage jersey searches at all-time highs and Victor Wembanyama at a career-high search volume. These are live demand signals, not forecasts, and for advertisers running search campaigns in apparel, retail, sports merchandise, or any category adjacent to fan behavior, the search volume shift is operational. Paid search strategies built around evergreen keywords are competing this week with a demand spike of documented scale.

Search Engine Roundtable reported on June 10 that Google is testing a new format for sitelinks within sponsored search listings: blue dotted underlines replacing the standard sitelink presentation. The test is cosmetic rather than functional, but visual changes to ad formats in SERPs affect click-through rates regardless of intent, and ad operations teams tracking performance anomalies this week should account for the possibility that format testing is a contributing variable.

Also from Search Engine Roundtable on June 10: a new study documented rejected Google review replies. Business owners who respond to customer reviews sometimes find their replies removed by Google without notification. The study analysed what types of replies are rejected, finding that Google deletes responses silently and that the business owner has no audit trail of what was removed or why. For local search practitioners managing review response programs at scale, the practical implication is that reply removal is more common than is generally assumed, and the absence of notification creates a quality assurance gap.

Schema.org added usage statistics to each schema type on June 10. The data is now visible on the Schema.org site directly. Author schema, for example, is deployed on over 10 million domains. Event schema is deployed on under 1 million. For practitioners implementing structured data or advising clients on schema priority, this is a live reference dataset for understanding relative adoption across the ecosystem.

DV360's Demand Gen API and new platform partners advance programmatic CTV

Adelaide added AU pre-bid targeting to Amazon DSP on June 10. Buyers can now screen Display, Online Video, and Streaming TV inventory by Adelaide's AU attention score before placing a bid, filtering out low-attention placements at the auction decision layer rather than in post-campaign analysis. AdExchanger reported the same day with additional technical detail: Amazon DSP is also integrating Adelaide's AU Quality Floor feature, which excludes inventory with AU scores in the bottom 10 percent. The Quality Floor also defaults to excluding made-for-advertising sites as classified by Jounce Media. The Quality Floor is exclusive to Amazon DSP for now; AU-based pre-bid targeting is available in other DSPs including The Trade Desk, Viant, Yahoo DSP, Adobe Advertising DSP, and Equativ.

Amazon Ads unified reporting reached general availability on June 10, consolidating campaign data across accounts in a single interface. Two legacy Ads Console reporting tools retire December 31, 2026.

Acxiom's InfoBase Geo-Based audience segments went live in The Trade Desk's data marketplace on June 10, covering the UK and Germany. Location-anchored consumer data from Acxiom's InfoBase product is now activatable programmatically within TTD campaigns in both markets.

Fubo signed a distribution deal with NBCUniversal on June 10 restoring NBC, Bravo, Telemundo, four NBC Sports regional sports networks, NBCSN, and new FAST channels to subscribers. The deal adds live sports and entertainment ad inventory to a platform that was missing it. MediaPost reported on June 10 that Paramount Skydance, in the process of absorbing Warner Bros. Discovery, is examining a plan to sell off Nickelodeon and Cartoon Network as a potential concession to appease California and New York state regulators preparing antitrust opposition to the deal. The kids cable asset disposal, if it proceeds, would reshape the premium kids advertising inventory landscape as those channels move to a new operator.

The Trade Desk named David Haddad, former president of Warner Bros. Games, to its board on June 10, adding media and entertainment expertise as the company rebuilds after a Nasdaq compliance notice issued following a period of stock price pressure in late 2025.

Minerva launches, first-party data startups accelerate, and the EU orders Meta to open WhatsApp

Minerva launched publicly on June 9 with $20 million in seed funding from The General Partnership, 8VC, Lingotto Innovation, Topology Ventures, and NBA Investments. The Brooklyn-based company was founded by Jackson Engles, Daniel Saedi, and Matthew Joseph, who worked at Lazard, Bridgewater, and Citadel respectively before building a platform targeting broken first-party data inside marketing teams. The product combines an Agentic Data Engineer for unification, a proprietary identity graph with over 1,000 consumer attributes, campaign creation and optimisation automation, and performance reporting. The OpenAI collaboration uses GPT-5.5, released April 23, 2026, across two product workstreams whose agents execute defined tasks without requiring machine learning expertise to supervise them. Approximately three dozen customers were already signed before the public launch. Sales, marketing, and CRM startups raised approximately $3.7 billion globally in early 2026, with agentic AI companies capturing a disproportionate share.

The European Commission ordered Meta on June 10 to restore free API access to WhatsApp for competing AI assistants, citing serious risk to competition in the AI market. The order requires third-party AI systems to access WhatsApp on the same technical terms as Meta's own AI products. It follows the Commission's earlier action under the Digital Markets Act and establishes a precedent for how dominant messaging platforms can structure access controls around AI integrations in regulated markets. The ruling lands on the same day as the Munich AI Overviews judgment, making June 10 a notable day in the calendar of European regulatory action against large AI-platform behaviours.

The Spend's most recent story: the SpaceX IPO that could move the entire market

The Spend's last-published item before the June 10 window was dated June 8: the SpaceX IPO pricing and listing timeline. SpaceX is set to price its shares Thursday and list Friday. The $75 billion offering is the largest in recent memory, and the story examines how the listing may trigger a chain reaction across global stock markets given the size of the offering relative to available liquidity. The Spend's framing is not that the IPO is good or bad for markets, but that the mechanics of a listing this large, at this moment, forces index funds and institutional allocators to either participate or explain to their stakeholders why they did not. That is a different kind of market pressure from a normal large IPO.

While not a programmatic advertising story, the capital market context matters in two ways for the ad tech sector. First, ad tech equity valuations, programmatic company financing conditions, and the cost of capital for businesses like The Trade Desk, Magnite, and LiveRamp all track broader equity market conditions. A $75 billion IPO that pulls capital from existing positions reshapes the investment environment for everything else listed. Second, SpaceX's Starlink business has significant implications for the global internet access landscape, and internet access expansion is the single most reliable predictor of addressable audience growth in programmatic advertising markets. The IPO transforms a closely held private company into a publicly accountable one with disclosed financials, which changes how much the industry knows about the economics of the infrastructure underlying digital connectivity globally.

The Spend's June 8 item on the AI bill that nobody budgeted for is also relevant context: token economics are reshaping how companies pay for AI, and the bills are coming in larger and less predictable than anticipated. For marketing teams that have moved AI tools from pilot to production, the infrastructure cost is becoming a line item that competes with media budget. That is a structural pressure on programmatic spending that will not be visible in platform revenue data but will show up in advertiser budget allocation decisions through the second half of 2026.

Also noted

  • June 10, 2026: APA and Edison Research surveys put 2025 US audiobook revenue at $2.43 billion, up 9 percent year over year, with 157 million active US listeners and over 750,000 active titles. YouTube piracy of audiobook content is rising, and willingness to purchase AI-narrated titles dropped sharply in the 2026 survey versus the prior year, a finding with direct relevance for audio content strategy and AI voice licensing.
  • June 10, 2026: Novi analyzed 10.7 million ChatGPT citations across more than 98,000 sources, finding that Reddit outranks Sephora and Allure when ChatGPT recommends beauty, skincare, and fragrance products. The data reveals a structural gap between where beauty brands publish controlled content and where AI systems find the content they trust enough to cite.
  • June 10, 2026: Semrush analyzed 89,000 LinkedIn URLs cited by AI search tools, mapping the content structure patterns, employee advocacy signals, and publication behaviors that correlate with AI citation rates on the professional network. The findings provide the clearest published picture yet of how professional network content earns AI visibility.
  • June 10, 2026: Kinsta launched free Bot Protection for all WordPress plans, giving site owners dashboard-level control over AI crawlers and automated traffic without server configuration. Apple's June 8 Applebot documentation update added AI training controls and nosnippet directives tied to how Siri processes publisher content for AI-generated responses. These two announcements on consecutive days reflect the same publisher-side pressure: the need to govern which AI systems access what content, on what terms, with what controls.
  • June 10, 2026: EMARKETER projects US AI ad spending will more than double to $68.25 billion by 2030, while placing ChatGPT's US chatbot advertising ceiling at $5 billion by the same year, with over 80 percent of total AI ad spend flowing beside AI-generated content rather than inside chatbot interfaces. The forecast frames the distinction between advertising adjacent to AI and advertising inside AI as the defining budget allocation question of the next four years.
  • June 10, 2026: Google DeepMind selected 15 European robotics startups for a three-month equity-free accelerator with access to Gemini robotics model infrastructure and direct DeepMind mentorship. Apple's iOS 27 announcement on June 10 introduced Tap to Share for contactless data exchange at retail checkout, Local Lists in Maps, and Visual Intelligence bill-splitting, none of which are ad products but each of which creates a behaviour surface relevant to local and retail campaign planning.