Two ways of building toward the same goal.
Full-funnel measurement is moving from concept to roadmap: broadcasters want to show not just who watched, but what audiences did next. That means uniting viewing, advertising and outcome data that usually sit apart to potentially develop a new standard for Reach and Performance (R&P).
Two initiatives show the routes broadcasters are taking.
TF1 is building it inside its own ecosystem. Its 2027 roadmap names full-funnel measurement a strategic pillar: the TF1 Efficiency Framework pairs a “Brand Simulator” for brand lift with in-store/on-site conversion tracking and cross-channel reporting, plus AI “Autopilot” agents joining targeting and trading from 2027 (VideoWeek, 10 September 2026). An earlier phase of Autopilot, already in production, cut a campaign-reporting job from four days to under five minutes for TF1PUB’s sales and analyst teams (Artefact, 9 December 2025).
Project Lantern is building it collaboratively across 3 broadcasters. Since September 2024, Channel 4, ITV and Sky, rival broadcasters, have worked with YouView to combine set-top-box data from around six million UK households with an independent outcomes panel. The first “Baby Lantern” results, at Vision 2025, found around 45% of TV-attributed outcomes were website visits, with the rest spread across search, social and in-app activity. As of March 2026, Lantern remains an ITV priority, working towards an “always-on” system (CSI Magazine, 26 March 2026; background via WARC and IBC).
The trade-offs
TF1’s route is fast: no consensus needed, already shipping. The cost is that its numbers live on TF1’s own terms which is good for TF1’s own campaigns, not comparable to a rival’s. Lantern buys the opposite: the use of numbers three competing broadcasters and YouView agree to mean the same thing, at the cost of two years to a first readout. Build fast on numbers you control, or slower on numbers the market can use.
The advantage neither has to build
There’s a third position that already has much of what TF1 and Lantern are building toward. Google’s advertising business made up around 68% of Alphabet’s revenue in Q2 2026 (ppc.land, 22 July 2026) — but most of what Google runs isn’t advertising at all: Search, Android, Chrome, Maps, Gmail, Cloud. Together they give Google a logged-in, cross-product view of users that feeds straight into YouTube and its ad platform. TF1 and Lantern are each assembling a full-funnel view from inside media; Google’s arrived largely as a by-product of businesses that aren’t media at all — a head start built into the architecture, not something a media owner can out-build.
GDPR is part of why the TF1/Lantern trade-off exists: combining data across separate legal entities means a consent chain and data-sharing agreement at every join — the friction that makes a Lantern-style build take years. Most of Google’s data sits inside Google, under its own terms, so that particular friction applies much less. No vendor changes that; it’s the backdrop every roadmap here competes against.
Where the difficulty sits
The hard part is rarely the dashboard. Audience data, advertising exposure and commercial outcomes sit in separate systems, using different identifiers and definitions. Building a consistent view requires agreement on what measures such as reach and attribution mean across those sources.
Streamhub supports individual media organisations and wider industry collaborations by providing the audience data architecture needed to connect audience, content, advertising and outcome data. It brings together different datasets, currencies and measurement approaches through transparent definitions, while allowing each organisation or consortium to retain control over its methodology and partnerships
This enables Streamhub to support proprietary measurement initiatives, shared industry frameworks and models that combine elements of both.
Rathish Radhakrishnan
Lead Product Owner