In Austria, the government taxes advertising. That makes it one of the few countries where ad spend is a matter of public record. The numbers are not flattering for traditional media: in 2018, Austrian media organisations collected roughly €2.2 billion in advertising revenue. 

Big Tech collected €862 million. By 2025, those figures had nearly inverted, media had fallen to €1.8 billion while Big Tech had surged to €2.7 billion.

Austria is not an outlier. It is a ledger.

The same dynamic plays out across every European market, in every advertiser category, at every budget level. Audiences, attention, and ad budgets are migrating toward platforms that offer what advertisers have always actually wanted: the ability to reach specific people, not just specific placements.

Programmatic and real-time bidding promised this in the mid-2010s but mostly delivered scale without precision. Mobile and social amplified reach while eroding CPMs. Then GDPR, Apple's App Tracking Transparency, and the slow death of the third-party cookie removed the mechanism that had made audience targeting possible in the first place.

For a decade, the gap between what Big Tech could offer advertisers and what everyone else could offer kept widening. Google and Meta had something publishers and ad networks did not: a persistent, consented, first-party relationship with hundreds of millions of users that let them build interest profiles, model intent, and activate campaigns with minimal waste.

That gap is now closable. A small number of organisations have figured out how to close it. The results are significant enough that the rest of the market is starting to pay attention.

What COPE built, and why it works

COPE Content Performance Group is Austria's largest digital ad network. It operates under STYRIA Media Group, which owns a portfolio of leading news brands, magazines, radio stations, and digital marketplaces. That publisher parentage matters for what comes next.

Presented at the INMA World Congress of News Media in Berlin in May 2026, COPE's Head of Research and Development, Maresa Wolkenstein, described a project that had been underway for several years: building what she called "the largest reach of addressable first-party data in Austria."

The architecture they assembled has four stages.

Consent, done properly. 
Users visiting any site in the COPE network encounter a TCF-compliant consent management platform. This is not a checkbox exercise, it is the foundation of everything that follows. Without explicit, informed, GDPR-compliant consent, none of the downstream data is legally usable for targeting. COPE treats consent as an asset, not a friction point.

A persistent cross-site identifier. 
Consented users receive a unique identifier via first.id, a European identity solution. This identifier follows the user across the COPE network, making it possible to recognise the same person reading news in the morning, checking sports scores at lunch, and browsing a marketplace in the evening, without relying on third-party cookies.

A user graph that builds interest profiles. 
User signals are consolidated into a graph database. Rather than simple page-level contextual signals, the system constructs a genuine interest profile, understanding what a user consistently engages with, what they are likely in-market for, and how they cluster with similar users. This is the layer that transforms raw behavioural data into targeting intelligence.

Audience activation at scale. 
Those profiles are turned into precisely defined audience segments. When a campaign runs across the COPE network, it activates those segments through Equativ (a European ad server) and first.id, delivering impressions to the right people across the network with what Wolkenstein described as "minimal wastage."

The full stack is built on European infrastructure, compliant with European law, and independent of any US platform. COPE framed this explicitly as a strategic advantage: an open environment rather than a walled garden, transparent rather than opaque, brand-safe rather than algorithmically unpredictable.

What the numbers showed

COPE ran a series of early test campaigns against this infrastructure. The results were compared against standard placements without audience data targeting.

In health: +200% CTR. In pharma supplements: +200% clicks. In luxury automotive: +11.5% CTR. In travel: +900% more inventory delivered to the intended audience. In pharma OTC formats, parallax ads delivered +62.5% CTR and siteBar formats +127.5% CTR.

The “philoro” case study, for a gold investment awareness campaign, illustrates how the pieces work together. COPE built a custom audience for philoro drawn from first-party behavioural data, enriched it with users identified through a "financial bad news" contextual theme (people reading about economic uncertainty), and ran two creative variants, one emotional, one rational, across premium formats. 

The campaign beat the Personal Finance benchmark by 34.84%, outperformed the DACH format benchmark by 21.86%, and delivered +162.5% dwell time versus benchmark.

Stephan Dangl, Head of Marketing at philoro, noted: "It was really important for us that our awareness campaign was based on reliable data, and we've rarely seen click-through rates like this, not even on finance sites."

One advertiser's response at the end of a campaign: "We'd like to move some of our performance budget over to you."

That last line matters more than any CTR figure. Performance budget, historically the domain of Google and Meta, is moving toward publisher-backed, first-party-data-driven inventory. The question is which networks are positioned to capture it.

The same bet, made on the other side of the world

COPE is not alone in making this move. News Corp Australia, presenting at the same INMA conference in May 2026, described a structurally identical approach built over several years.

Their backend infrastructure centres on a Customer Data Platform (CDP), which connects first-party signals from a portfolio of owned brands, news.com.au, The Australian, Herald Sun, Daily Telegraph, taste.com.au, Escape, Vogue, and others, into a single customer view. 

Layer on top of that identity signals from several different identity providers (services with authenticated users), plus enrichment signals from national and global data providers, and the result is an audience intelligence capability that rivals anything available on the open market.

The front-end product gives media buyers a self-serve interface to explore audience insights, plan campaigns, activate buys, and measure outcomes, the four steps that advertisers expect from any modern platform. It replaces the traditional agency phone call with a data-driven planning environment built on first-party intelligence.

News Corp's roadmap goes further. They are already working on unifying data sets to expand addressable audience scale, developing consistent data schemas across their CRM, ad server, booking system, and inventory management platforms, and building AI agents that can interact with those schemas, including external-facing agents using emerging agentic protocols (AdCp) to automate trading.

Two organisations, two continents, the same conclusion: first-party data infrastructure is not a feature. It is the business.

The gap that ad networks without inventory face

COPE and News Corp built this infrastructure from a position of strength, they owned the inventory, they owned the audience relationships, and they had the first-party signals flowing in from their own properties. The challenge they solved was making that data addressable and activatable at scale.

Ad networks that operate without their own media inventory face a different version of the same problem, and in some ways, a harder one.

They sit in the middle of the value chain: they sell into publisher inventory on one side, and represent advertiser budgets on the other. For years, that position worked because they could offer reach across a broad portfolio of publishers, and targeting was handled by third-party data and cookies. Both of those advantages are eroding. Third-party cookies are either gone or going. 

Third-party data is increasingly restricted by regulation and scrutinised by advertisers. And the publishers worth buying through are rapidly developing their own first-party data capabilities, which means they will increasingly prefer to work directly with buyers who can match into their audience data, rather than through intermediaries who cannot.

The advertiser expectation is now set by what Big Tech offers: reach, frequency, audience targeting based on real intent, cross-device consistency, and outcome measurement. An ad network that cannot participate in that value exchange does not get replaced by a better ad network. It gets bypassed.

The question is not whether ad networks need a first-party data infrastructure. It is how they build one without owning the underlying inventory.

What that infrastructure looks like for a network

The COPE model provides a useful blueprint, adapted for networks that aggregate publisher inventory rather than own it.

The foundation is the same: consent-compliant user identification across the publisher sites within the network. Rather than each publisher operating its own disconnected data silo, a shared identity layer, using solutions like first.id or equivalent, makes it possible to recognise users across multiple publisher environments with appropriate consent. The network then becomes the aggregator of those signals rather than a single publisher doing it alone.

On top of that foundation, behavioral signals from across the publisher network feed into audience construction. A user who reads automotive content on one publisher site, financial news on another, and travel content on a third is building a coherent interest profile across the network, one that is more valuable than any single publisher could construct on their own. 

The network, sitting across all of those publishers, is the natural home for that aggregated intelligence.

Those audiences then become the asset the network sells to advertisers. Not placements. Not CPMs against a contextual guess. Specific people, identified by real behavior, reachable across a compliant cross-publisher network, measurable against outcomes the advertiser actually cares about.

This is what COPE built for its own network. It is what News Corp built for theirs. It is increasingly what advertisers expect from any network that wants to compete for ad budgets.

Why the timing matters

COPE's own post-campaign learnings are candid about what remains difficult. The market is still heavily focused on CPM and impressions rather than outcomes. 

Sales teams have to work harder to explain the difference between new first-party-data targeting and the "old" targeting that buyers have grown accustomed to. Prices need to go up to reflect the genuine value being delivered.

But these are adoption curve problems, not structural ones. They describe a market in transition, not a market that has rejected the direction of travel. And transitions reward early movers disproportionately, the networks that can already demonstrate first-party audience targeting performance, when a sophisticated advertiser asks for it, will capture the budgets that competitors cannot even bid for.

The leading networks in Europe are already asking these questions. Some are making the investment. The ones that figure it out in the next twelve to eighteen months will define what premium network buying looks like for the rest of the decade.

COPE already has its answer. News Corp has its answer. The market will catch up to them. The question worth sitting with is: where will your network be when it does?

Closing comments

Samhub was born from Swedish publishers with the aim to help publishers and media collaborate around first-party data. The goal was to build “data muscles” to compete for ad budgets with the global players. 

If we do a side by side comparison of what COPE, News Corp and Samhub does the similarity is striking. To me it shows that what we started in 2018 is today more relevant than ever.

As we see the sell-side accelerate into first-party data there is a need for cooperation and unification across the programmatic ecosystem. And as media starts implementing first-party data infrastructure (slowly but surely) the choice for ad networks is to either tap into the publishers first-party data or integrate/build their own data infrastructure from scratch, just like COPE has done.