I’ve been building a case this year in Berlin, Delhi, Vail and Madrid that I pulled together in full last week in my INMA Advertising Webinar “Advertising Transformation: Everything You Need to Know in One Hour.” INMA’s Paula Felps wrote a fantastic recap, and the recording is available to members at the link above. 

You’ve heard bits and pieces of this the last few months, but I’m going to recap it in full here so you have it in one place.

The diagnosis

Eighteen months of immersion in markets around the world since we relaunched the Advertising Initiative has crystallized one thing above all else: Everything that is happening to you is happening everywhere to everyone else. The only differences are speed, timing, and impact, and those depend on how much of your revenue is print versus digital and advertising versus reader revenue.

Traffic is falling (40%-50% is the number I hear most), creators are becoming the main way young people get information, mobile vertical video is the medium they prefer, and AI is across all of it, including the fact that more than half of all Web traffic is no longer human. 

The business built on traffic is dead. Ad spend on news is down 33% since 2019 (WARC), while global advertising will pass US$1.6 trillion by 2030 and Google, Meta and Amazon take 60%-70% of it, as much as 90% in some markets.

Buyer demand is up for almost everything except what publishers sell: creators about +24%; CTV, social video, retail media, and digital audio and podcasts (about +10%) — the one bright spot in the traditional publisher portfolio. Down: display, open-Web CPMs and print.

The paradox

The super users I wrote about two weeks ago, the 1% who drive most of the revenue, matter more every quarter because the least valuable users are the ones leaving the fastest. 

The New York Times found the same thing from the advertising side, as Valerio Poce presented in my first-party data master class last year: The most engaged readers are the most valuable consumers across categories and the most receptive to advertising. 

And yet we’ve built a business that trades subscription value for advertising: Pay us enough and we’ll turn the terrible ads off. It’s strategically incoherent but completely explainable when you acknowledge that the pop-up-riddled, ad-dense, programmatic-slop version of the ad business is so hostile to the reader’s user experience that it only makes sense to turn it off for the people we care about the most.

Meanwhile, the buy side is scoring inventory on ad density, domain authority, and content quality. And as supply falls with traffic, demand is chasing fewer good impressions. 

In July, a Digiday reporter working on Ozone’s Q2 benchmarking told me U.K. publisher ad supply had fallen about 40% and higher eCPMs made up most of the difference. Fifteen publishers in South Asia told me the same thing later the same month. The ANA confirms that in 2024, open programmatic was the majority of what was traded, and two years later private marketplaces are. And The New York Times — with lower ad density, majority non-traditional display, and a concerted focus on the quality of the user experience — has had 10 straight quarters of digital advertising growth, six of them double-digit.

So, three imperatives:

• Stop turning ads off for the most valuable people. 

• Go after the trillion-plus U.S. dollars publishers don’t get, not just the US$30-US$50 billion they do. 

• And make the ads worthy of the journalism, the audience, and the brand.

The playbook

Rethink inventory: The New York Times Flex Suite, larger formats that perform better and cost more. Russmedia’s micro apps as ad inventory, vibe-coded and embedded in editorial. If we’re free from the banner, what do AI, structured data, liquid content, and a real partnership with clients let you build and sell that you couldn’t before?

Solutions and outcomes, not reach: Mediahuis’ In-Market finds people actively shopping for a product. Hearst’s Aura IQ and the Times’ Brand Match use generative AI to build a custom audience for every campaign from first-party data and content signals — and both consistently beat benchmarks. Bonnier News, with 60% daily reach in Sweden, paired with Visual Art, which controls 85%-90% of the country’s in-store digital screens, to sell one full-funnel campaign. 

Bake advertising in: The transformation we need is not possible if advertising only shows up after the product is built. That drives interruptive formats and awkward placements. Events work because editorial, marketing, product, and advertising are designed together.

Structure for transformation: The publishers getting this right have given advertising a seat at the table with integrated structures like Bonnier News and shared KPIs like News Corp’s engaged reach. Move inventory from open programmatic toward private marketplaces and direct, where CPMs are higher and fewer ad tech vendors take a cut. Build measurement beyond reach, frequency, and CTR because nobody will tell your story better than you. And expect the team to invert: from 80% sellers and 20% support to a smaller group of relationship sellers backed by ad ops, data, studio and analysts.

What’s defensible

The market wants outcomes. 

Since COVID, performance has taken about 70% of budgets, inverting the old 60/40 brand-to-performance split, and that suits platforms that sell hard outcomes (sales, leads, conversions) while publishers mostly contribute awareness, brand lift, and intent. 

There are early signs of a turn: In ISBA’s 2026 U.K. media budgets survey, 37% of advertisers planned to increase brand investment versus 14% prioritising performance. And the aggregated brand lift studies I have seen say publishers are best not at awareness but at shifting consideration, the hardest thing to move.

What’s defensible is some version of brand, trust, and relationships, sold three ways: 

• Attention: the quality of the time people spend with the journalism.

• Influence: your credibility with your audience and a brand’s chance to be part of it.

• Intent: evidence that you can deliver a motivated audience. 

AI will be the ultimate test of whether I’m right or wrong. Agents will automate habit and compress research, so habit categories like grocery and CPG are less defensible. Meanwhile advice categories (finance, B2B, healthcare) and desire categories (luxury, travel, fashion, beauty) are where publishers can defend and grow spend if they keep creating intent and can prove it. 

From inventory to brand partnerships

What came together for me in Spain is that we don’t love the banner ad version of the ad business, and we never have. 

Nobody sounds proud of their programmatic display, and banners don’t win Titanium Cannes Lions. When people talk about brand partnerships executed through events, creators, branded content, and video, they take a different tone because partnerships better balance the ambition of the newsroom, product, brand, and advertising teams. 

The next audience is there, too: 90% of Spil’s young followers are open to branded content. Young audiences are not anti-advertising. They are anti-bad advertising, and so is everyone else.

So the mental model is a mirror of print to digital: From a strategic perspective, start the transition from the ad inventory business to the brand partnerships business. Don’t actually stop selling inventory, and maybe you never will! Protect the old business as long as you can, maximise it, and use it to fund the new one — but see the future clearly.

One more time, as clearly as I can deliver it:

1. Build for what’s defensible, not what’s dead.

2. Make the ads worthy of the journalism, the audience, and the brand.

3. And just like print-to-digital, transition from the old business (ad inventory), to the new one — brand partnerships.

If you’d like to subscribe to my bi-weekly newsletter, INMA members can do so here.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .