Nic Newman opened his Media Party session with a heatwave warning. During the worst of the summer, he had put a handful of questions to Claude. Back came a sourced Q&A, a temperature chart, an energy-mix chart by country and a list of the hottest spots, then the same again in Spanish. “It assembles this information very quickly. Assembles it essentially as a service,” the Reuters Institute’s senior research associate told the room. “From a publisher point of view, of course, this looks more like an existential threat.”

Later in the morning Dmitry Shishkin opened his session with a sentence. “It doesn’t really matter what you do with AI. If your content is crap to start with, then you are not going to succeed.”

Though the programme listed them as separate sessions, the two amounted to a single argument delivered in halves. Newman had the stats: where audiences and referrals have gone, and where publishers say they are placing their bets. Shishkin, the former BBC editor behind the user needs model, had the case that those bets cannot be paid out by newsrooms whose editorial, product and data systems do not talk to each other.

The referral era ended between two Novembers

Newman’s opening chart, aggregated across the 48 countries of the Digital News Report 2026, showed social media overtaking television as the most important source of news for the first time. He was more interested in a line lower down. “Access to news websites and apps are down 12 percentage points since 2021, as people find new and more convenient ways to access the news.” Across countries, 10 per cent now regularly use chatbots such as ChatGPT or Gemini for news.

The supply-side data was worse. Publishers surveyed for the Institute’s trends report at the start of the year expect to lose around half their traffic over the next five years as content is absorbed into AI interfaces. Chartbeat data covering some 2,500 sites shows Google referrals falling by about a third between November 2024 and November 2025, a decline Newman said was not universal and not wholly down to AI Overviews and AI Mode. “The key story is that publishers won’t be able to rely on referral traffic in the way that they have done in the past.”

Publishers have chosen distinctiveness. Shishkin asked whether they can measure it

Asked where they would put more and less effort, publishers gave Newman a consistent answer: more into original investigations, on-the-ground reporting, contextual framing, narrative stories and face-to-face events; less into Q&A, general news and some service journalism, which AI already does well enough.

Shishkin’s session took that strategy at its word and asked what it requires. His opening question to any client is whether they are “unique enough to actually justify people spending time with you”, and then he asks what would happen to their audience’s lives if the organisation disappeared tomorrow.

Every topic a publisher covers, he argued, splits three ways: commodity content everyone has, and which can be automated, expert content only that newsroom can produce, and services built around the topic so that readers can do something with it. “Content by itself is not going to be enough for people to keep coming back.” Most organisations cannot say how their own output divides. “If I were to ask a lot of media, they will not even be able to say what is the balance between our generalist content and expert content, for example. Nobody is tracking that.”

Liquid content needs solid building blocks

Newman’s examples of promising AI work all involved content coming apart and reassembling. Aftonbladet’s front-page chatbot prompts now draw 250,000 questions a week. The BBC, which has text for every Premier League club but could never staff a studio for each, is testing AI-generated daily audio briefings per club in the right regional accent. VG in Norway has built an experimental app for younger readers from its existing content, with no articles in it. “There are no editors, so all there is a system and monitoring after publication.”

The same survey found the industry unmoved by AI’s commercial effect. AI is embedded in research, headlines and workflow at almost every publisher, but heads of AI still decline to call it transformational. The vast majority said no jobs had been cut.

Shishkin’s diagnosis of that gap is structural. Newsrooms run newsletters, social, analytics, video and a CMS, and none of the data connects. “In order to be successful in the AI world, you need to create a shared taxonomy across everything that you do, you need to have one version of truth.”

In practice that means tagging every piece of content three ways. First, a detailed topic, not a single word like politics. Second, a format: interview, feature, write-up or Q&A. (Audio, video and text, in his scheme, are content types, not formats.) Third, the user need it serves. Work with smartocto has shown a personal finance story performing three times better as a listicle than as a write-up, a difference invisible unless format is tracked.

That tagging is what makes Newman’s liquid content intentional. Turning an updating story into an educating one is easy, because the archive supplies the material. Turning it into an inspiring one requires assets gathered on the ground, so the commissioning meeting has to know the permutations before the reporter leaves the building. “It’s not about creating better journalism anymore. It’s also about building systems that deliver the right journalism to the right person in the right form at the right moment.”

The agent will read paragraph five, if anyone tells it where to look

Newman’s next stage is agents working in the background on the reader’s behalf, and publishers expect them to do more damage than chatbots. Shishkin’s version is blunter. “You can be running a really important newsroom where important journalism is being done, and then everything will be completely flattened by AI agents that will come and pick something from you.”

His answer, in development since May, is an agent readiness layer: eight signals covering recency, provenance, attribution, sourcing, clarity and structured claims, so that content is retrievable, verifiable and comparable. His illustration was a ten-paragraph article visited by an agent comparing five sources. “Instead of them reading the whole article, they land on your page and you say, actually, your task means that paragraph number five is the most important paragraph for you to take.” At Monday’s workshop, participants scored a BBC News article and a Guardian article on the same story against the eight signals and agreed the Guardian did better. He offered the worksheet to anyone who asked. “Nobody else will solve it for us.”

Trust has moved to individuals, and Goalhanger has built a company on it

Newman’s second squeeze runs through people. Social media has stopped being social: a Financial Times chart shows only 17% of a Facebook feed coming from people the user knows. “These are content distribution networks designed to engage,” and the engagement is going to video networks and the creators on them. Respondents as a whole rate creators as more entertaining and easier to understand than journalists, and as less knowledgeable and less trustworthy. The quarter who actually use them rate creators better on every measure, trust included. “You see what you get. Whereas with news brands, there’s potentially hidden agendas.”

Publisher responses so far amount to making journalists more like creators, from Wired’s platform personalities to straight-to-camera video on the New York Times site. The model Newman singled out is Goalhanger, no longer in his description a podcast company. “It’s really about shows and talent.” Around the show sit audio, video, text, newsletters, social and increasingly television; around those sit membership, licensing, advertising, merchandise and ticket sales. Its South Bank event the weekend before the conference sold out.

In Shishkin’s terms that is a service built around content. It is also the one case where the taxonomy question barely arises: the relationship belongs to a person, and the format is the show.

On Monday, Media Party, the journalism and technology conference founded in Buenos Aires, opened its first European edition at CaixaForum Barcelona, with close to 600 registrations and a hackathon scheduled for day three. Marc Melillas, the Generalitat’s Director General of Media, used his welcome to note where the money goes: companies commit millions to AI while almost nothing goes into the ethics of how democracies use it. Mariano Blejman, the founder, listed what the format has produced, including GPTZero, presented by Edward Tian at the 2023 Chicago edition; it went on to raise around $15m. Media Party, Blejman said, is evaluating opening a space in Barcelona. The three keynotes that followed all turned on trust, and each speaker meant something different by it.

Verifying the output is the real cost

Teresa Mondría Terol of NPR worked inside the New York Times newsroom on the Epstein files and used that investigation as her anatomy lesson. Her test before any AI project has two parts: which stories become possible when time is no longer the limit, and where checking the output will cost more than the tool saves. “In the end it generates more information than you had at the start, and that doesn’t help us.”

Reconstructing the relationship between Steve Bannon and Jeffrey Epstein in the months before Epstein’s death meant finding their exchanges inside millions of documents, from flight records to iMessages. The two did not always call each other Steve and Jeffrey; initials and nicknames recur. An AI agent working in natural language found the first traces. Machine learning pulled every page on which either addressed the other, and generative AI extracted dates from messages in inconsistent formats. Every message was then verified one by one.

The same team described images and video frames to build a semantic search, so reporters could look for Epstein’s New Mexico ranch in words, and built agents that let journalists new to the story query the documents with citations back to the originals. For the second anniversary of 7 October, AI extracted everyone from or living in Gaza whom the paper had spoken to over two years; verification meant reading each source document in full and re-running the instructions to catch omissions.

The methods are not confined to investigations. The New York Times runs hundreds of hours of manosphere podcast audio through models to produce five-minute daily digests, so the reporters on that shift can follow the discourse without listening to all of it; the Wall Street Journal, she said, does the same. A local media group she cited transcribes and summarises recordings of public meetings no reporter attended. An open-source plugin now let’s newsrooms run document-by-document classification inside Google Sheets; she uses it herself, and at NPR is building internal search engines tailored to the production processes of individual podcasts. Her advice for newsrooms starting out: use it early in an investigation, when the job is locating the metal in the haystack, and in internal processes where mistakes surface sooner.

3Cat is chasing trust to wherever it settles next

Judith de Argila, Director of Digital Audience Development at 3Cat, Catalonia’s public broadcaster, described a news transformation begun from a position of strength. 3Cat leads television news in Catalonia. The result is 3Catinfo, a news brand built on integrating television, radio and digital production.

She named three problems.

Visibility: algorithms set the agenda, and the live questions include whether 3Cat’s content surfaces in ChatGPT and whether its app sits in Samsung’s marketplace.

Disconnection: journalism’s taste for nuance and data reads to many people as elitist. Trust: the hierarchy in which governments and media set the agenda has gone.

The responses: content made natively for social platforms; WhatsApp, Telegram and Instagram channels outside algorithmic feeds; distribution deals sought with Spotify and Samsung; an app rebuilt around scroll and vertical video; a verification unit whose head appears on bulletins explaining how stories were checked, with capsules pushed through WhatsApp to counter falsehoods on the same channel. The media, she added, are lobbying at European level for AI companies to prioritise their sources in answers. The 3Catinfo Instagram profile now reaches around three million people with news content alone.

The broader claim was that trust is being deposited elsewhere, with independent journalists, creators and individuals, and the next question is what happens when people ask ChatGPT, Gemini or Claude instead of a journalist.

PRISA bet on patience over speed

Olalla Novoa Ojea, PRISA Media’s Director of Artificial Intelligence, described a year spent on governance. When the work began, the group already had more than 30 AI projects catalogued, some in production, and a choice. “We could carry on doing pilot after pilot,” she said, “or pause momentarily and build the foundations that would let us run faster afterwards.” A group-level responsible-use policy, approved in December 2024 to follow the EU AI Act, went to every inbox; she doubted many read to the end.

An AI Oversight Committee meets every two weeks, with legal, compliance, privacy, technology, AI, sustainability and HR at the table, and reviews use cases from newsrooms and non-editorial teams alike. An executive committee led by the CEO meets every two months to set strategy and take escalations. Every proposal answers three questions: what problem, how, and why, the last covering alignment with editorial principles and audiences. Tools and use cases are approved as a pair; change either and the review restarts. Nothing reaches the committee without a manager’s budget. “No money, no project.” She had expected the fortnightly cycle to become a bottleneck. “Quite the opposite: it runs like clockwork.”

The instruments are deliberately plain. A use-case sheet, adapted from a technology template to a journalistic setting, makes teams write down what, how and why before they ask. An intranet AI hub, still in beta, holds the policies, training on demand, a prompt library and a community of practice where staff post the problems they hit. Monthly office hours and two ambassador networks carry questions back to the AI team.

A year in more than 1,000 staff trained, though that is 60% of those invited; 21 use cases approved; eight projects live or in development. Shadow IT persists, and PRISA chose training over blocking. Agent governance is the 2027 problem, with a first agent sheet drafted, alongside vibe coding by staff who have never programmed, which has to be squared with the group’s secure software protocol. Her close named the asset all this protects: 50 years of El País, more than a century of Cadena SER, more than 75 years of Caracol. Competitive advantage, she said, will not go to whoever deploys the newest model. “Trust is built with governance.”

PRISA’s oversight committee meets every fortnight. The audience de Argila described is moving its trust to creators and chatbots on a shorter cycle than that. The gap between those two clocks is where the rest of the programme sits.

Readly has signed a distribution deal with EE that places its catalogue of more than 5,000 magazines and newspapers inside EE One Up, the rewards programme for EE One customers who take broadband and mobile from the operator together. EE puts the programme’s reach at over a million households.

Those customers get what the two companies call an extended period of free access to the Readly app. No duration is given. The catalogue includes The Guardian, Empire, Stuff, Edge, Women’s Health, Radio Times and Auto Express, covering news, entertainment, technology, gaming, health and motoring.

The deal is the first significant UK announcement since Cafeyn completed its acquisition of Readly’s business outside the Nordics in April. That transaction left the French group with combined revenues of nearly €100 million across 15 markets, 2.5 million users, and more than 5,200 publications from 1,100 partner publishers, and Cafeyn said at the time that it wanted to consolidate its position in key markets, notably the United Kingdom.

A retention play for EE, a UK foothold for Cafeyn

For EE, the arrangement is retention. One Up is where the operator parks the third-party perks meant to stop those converged households shopping around. Journalism joins that list, and its job is to cut churn.

Cafeyn’s interest is distribution. The group describes a hybrid B2C and B2B2C model: direct subscriptions on one side and, on the other, wholesale deals in which a telecom operator or digital platform pays to hand access to its own customers. It has built such partnerships across Europe, and the EE deal applies the pattern to Britain. Cafeyn’s pitch to operators is engagement and loyalty; to publishers, it’s that the wholesale fee enables it to reach readers they could not reach alone.

“Partnering with EE, one of the UK’s leading consumer brands and subscription platforms, is an important milestone for Readly and Cafeyn Group. It also reflects the growing role trusted editorial content can play in helping telecom operators create deeper customer engagement. Together, we aim to make high-quality journalism more accessible to millions of UK consumers through a seamless and highly valued experience.”

Laurent Kayser, CEO at Cafeyn Group

The publisher is paid for a reader it never meets

Several numbers are missing from the release. There is no figure for what EE is paying, no split for publishers, and no indication of what happens to One Up customers once the free period ends: whether they are asked to pay Readly’s standard rate, whether EE funds continued access, or whether the offer lapses. No conversion expectation is set.

EE holds the customer and the billing relationship. Readly holds the app, the reading data and the personalisation layer. The publisher holds a share of an undisclosed fee and a title on a shelf beside 5,000 others.

A million households are reached. Whether any of it survives the end of the free period as paid reading is the number nobody has offered.

A new burst of the ASA public awareness campaign went live on Wednesday across television, digital, out-of-home, radio, cinema, press and magazines. It runs for eight weeks, and none of the media was paid for.

Thirty-one organisations have donated the inventory, among them Channel 4, ITV, Sky, Google, Meta, TikTok, Spotify, Reach, News UK, Telegraph Media Group, Mail Metro Media and Ocean Outdoor. Leith was the creative agency that worked pro bono. WPP Media’s EssenceMediacom donated the planning. The Advertising Association says many media owners have significantly increased what they give.

Ten points of trust, and a recovery the AA itself calls fragile

Since the campaign began in 2022, public trust in advertising has risen from 30% to 40%, a five-year high, according to ASA research from October 2025. The industry’s supporting statistic is the one that keeps the donations coming: people who have seen or heard the campaign ads are twice as likely to trust advertising as those who have not. Previous figures from the Credos Trust Tracker, 2025, indicate that 25% of consumers distrust advertising, down from 29% and 33% of consumers who neither trust nor distrust advertising.

The AA’s language about the recovery being fragile is not modesty. It is an accurate description of a ten-point gain that took four years and could be given back faster than that.

The two new names say something about where regulation is least visible

Netflix will run the 30-second advert on its ad-supported plan. Waitrose will carry a bespoke ASA advert in Your Week In Food, the newspaper handed out in its stores. Both are additions worth more than their reach.

Netflix’s ad tier is a young inventory pool populated by viewers with no habit of thinking about who polices what appears between programmes, and streaming is precisely the environment where the assumption of regulation is weakest. Waitrose puts the regulator’s name in print in-store and within a few feet of the purchase decision.

Returning brand partners Churchill, Compare the Market, Marmite and Tesco have again lent their straplines to the campaign creative. Aleksandr the meerkat features in the updated television advert.

“Our purpose is to help make great financial decision-making a breeze for everyone, so customers are at the heart of everything we do at Compare the Market. We’re delighted to lend our much-loved meerkat Aleksandr to the ASA’s latest campaign, helping to deliver the important message that advertising is a regulated industry.”

Tom Wallis, chief customer officer of Compare the Market

The regulator being advertised is not the one doing most of the work

In 2025, the ASA resolved over 40,000 complaints about 25,397 ads, and 22,383 ads were amended or removed. Its AI-powered Active Ad Monitoring system processed 60 million ads. The shift is deliberate and documented in the ASA’s own annual reports: in 2012, 5% of its regulatory resources went to proactive work; by 2024, proactive cases accounted for 59% of its casework.

The complaints process is what the campaign asks the public to believe in, and it is a rounding error against the volume the machine now handles. The ASA has quietly become a monitoring operation with an attached complaints desk. Nothing in the creative reflects that, and there is no obvious reason it should. Reassurance and accuracy are pulling in slightly different directions.

“Advertising is part of everyday life, so it’s important people know we’re working to make sure ads are legal, decent, honest and truthful. This campaign plays a vital role in raising awareness of the ASA and our role in ensuring ads are responsible, which in turn helps build trust in the advertising people see. We’re hugely grateful to the brands and media owners supporting the campaign again this year and particularly pleased to welcome some new partners on board.”

Miles Lockwood, director of complaints and investigations at the ASA

Action 1 exists because the second century opens under review

The campaign is the first item on the Trust Action Plan the AA launched in Westminster in July, which called on the industry to scale it up. Scaling up a campaign with no budget means asking thirty-one media owners for more, which is what appears to have happened.

The AA has spent its centenary year arguing that self-regulation works, to a government whose instincts on gambling and food advertising run the other way, and in the same month as ASA rulings on the HFSS brand exemption drew a hundred-organisation backlash. A campaign nominally aimed at the public is being watched by a much smaller audience in Westminster.

Ten points in four years, on borrowed airtime, is the industry’s evidence that it can be trusted with its own regulation. It is also all the evidence it has.

IAB UK’s recently published Compass report on digital out-of-home forecasts that programmatic trading will account for 24% of all OOH spend by 2031. More than three-quarters of spend will still be traded outside programmatic channels, a decade and a half after programmatic reached outdoor. To put this into context, over 90% of display advertising is sold via programmatic channels.

Elizabeth Lane, who runs IAB UK’s insight programme and oversaw the report, treats the figure as a waypoint. She responded to written questions, declining others, including on data privacy and the forecast’s downside scenario.

“Programmatic DOOH is likely to continue rising beyond our current forecasts of 2031, so the 24% quoted in the report is very unlikely to be the ceiling,” she said.

“PrDOOH doesn’t have a structural ceiling per se, but as our Compass report explains, DOOH is unique for its premium environment and, in many cases, guaranteed audiences. As such there will always be great opportunities for brands to work directly with media owners, and it will likely mean some inventory will be traded in a non-programmatic way. However, we expect that 24% to continue rising well into the 2030s.”

The industry has a capacity problem: it can’t just build more screens

Programmatic has turned online display into a commodity and the yields have been reduced accordingly, so the question is why would outdoor fare any better? The answer is supply, says Lane.

“We don’t expect programmatic DOOH to show a similar trajectory to online display due to one factor, the scarcity of screens available. Unlike other ad formats, DOOH cannot create another space to show an ad, which helps protect [DOOH] as a premium environment. And this too helps DOOH remain a rich media, brand-safe ecosystem for advertisers and consumers alike.”

Elizabeth Lane, Head of Insights at IAB UK

Display yields fell because supply could be manufactured faster than demand grew. An outdoor screen needs planning consent, power, capital and a site somebody already owns. The digital estate has grown significantly, from 14,690 frames in 2018 to 41,517 in 2025, according to the report. Each one is an object bolted to something.

Asked how much programmatic DOOH goes through private marketplaces rather than the open exchange, Lane said the IAB UK holds no figure: “We don’t have data on the amount of prDOOH that is traded via the open exchange vs private marketplaces.”

Display’s prices collapsed on the open exchange, the end of the market where any buyer can bid on any impression. Outdoor’s premium survives in private deals between named buyers and sellers. A drift towards the open exchange would show up in prices well before anyone had the trading data to account for it.

The report’s recommendation for measurement collaboration is addressed to no one in particular

Route, the industry’s jointly owned audience currency, has counted outdoor’s audiences for years, yet the report still calls for greater collaboration on measurement. Asked why the currency has not settled the standardisation question, and who should own the standard among Route, Outsmart, the IAB and the SSPs, Lane answered in general terms rather than naming one body.

“Standardised measurement of digital advertising is an industry-wide challenge, and one that DOOH shares alongside other ad formats. There are many different trade bodies and JICs playing different roles within this, whether that’s audience currency, ad visibility, or outcomes measurement,” she said.

“The IAB’s view is always that openness and transparency across all of us will help move the industry towards the ultimate goal: consistent cross-media owner measurement and attribution for DOOH.”

A measurement standard is a competitive asset for whoever ends up owning it, and every trade body that would have to concede ground is funded by members with a position to protect, which is a reasonable explanation for why openness and transparency are the operative words and not, say, adoption or deadline. IAB UK set up a Measurement Advisory Board in January to work on this.

Only 4% of IAB UK members would call themselves agent-first

The report argues that concentrated ownership leaves DOOH well placed to adopt agent-to-agent trading standards such as IAB Tech Lab’s AAMP. IAB UK’s own research finds that 60% of the industry cite transparency, brand safety, and data security as significant concerns about agentic AI. Asked who gets disintermediated when agents plan and buy against live triggers, the OOH specialist agency or the DSP, Lane rejected the premise.

“I don’t think it’s really a case of one or the other being disintermediated, it’s more that both take on different roles in an agentic system,” she said. “Our State of AI in Advertising research shows most of the industry is still experimenting with agents rather than being fully agent-first, only 4% of IAB UK members would put themselves in that camp, so very little is being disintermediated in DOOH in the near-to-medium term. Advertisers still have real concerns around transparency and brand safety, and governance is something the industry is actively seeking guidance on, which is exactly why that specialist expertise is needed more than ever.”

On timing, she offered no date. “The first thing to acknowledge is that ‘agentic’ is a spectrum rather than one set way of transacting. We’re already seeing agents used across parts of the buying process, though these still involve a lot of human oversight, the expectation is that this decreases over time.”

Outdoor is not where the tests are running. “Agentic transactions in media is a truly nascent area of advertising, we’ve only seen test campaigns so far emerge in the UK in ‘traditional’ display advertising. It’s possible we’ll see some agentic transactions for DOOH campaigns soon, but it’s impossible to put an exact date on when.”

AAMP 2.3, she added, is “a framework that evolves today’s programmatic infrastructure into one where agents can trade and execute advertising while remaining interoperable and governed. The answer about what’s realistic is less about the technical capabilities, and more about addressing more fundamental barriers such as trust and where an agentic transaction delivers added value.”

You can download the report here.

“My mum was my school dinner lady. I had no industry connections.” Rachel Duffy grew up in Middlesbrough, where the industry felt out of reach; even Manchester and MediaCityUK seemed remote. Everyone she met on the way in seemed to have found work experience through someone they knew, or an uncle at a newspaper. She had a history and international relations degree, a university paper, some freelance social media work, and a plan.

On paper, the route she took looks clean: an NCTJ (the National Council for the Training of Journalists qualification) at News Associates, writing for the Londoners, then a job as Social Editor at The Telegraph found on LinkedIn. “It’s not something I advertise, but I applied for about ten different grad schemes and entry-level roles at most of the nationals. I didn’t get an interview for one.” Beyond the grad schemes were 50-plus applications to anything journalism-adjacent, most of which went unanswered. Some answered eventually.

“There were a lot of jobs that I heard back from many months after applying, by which point I had already been at the Telegraph for a while.” That, she says, “felt like salt in the wound”.

Rachel Duffy presents at the Reddit HQ

Duffy is now Head of Community at The Times and Sunday Times. The Young Journalist Network, which she founded and still runs, exists because of the gap between how simple her route looks in hindsight and how opaque it felt at the time. It operates under the founding principle that “Access and opportunity, not luck,” will sustain the next generation of journalism. “I didn’t know what qualifications were needed or what a good way in looked like. I would have loved to have a conversation with someone who was in the industry at that age. YJN tries to alleviate some of those stresses and unanswered questions, pairing young talent with someone that they can have those conversations with.”

Getting in is one problem; staying in is another

YJN runs in two distinct parts. The first serves undergraduate and postgraduate students trying to get in: CV and cover letter reviews, grad scheme application support and plain answers to questions the industry assumes everyone already understands. “People are throwing around these terms like NCTJ. So many people don’t know what they actually are or exactly how helpful they are.” Should I do an NCTJ? Should I go to City? The network has partnered with universities to field exactly those questions.

The second part is retention: monthly events for people already in the industry, mostly under 35, a generation that arrived after the post-work pub culture that once did the industry’s informal networking has thinned out. “It’s a very gruelling industry,” she says. “The shifts are brutal.”

When Reddit became the platform that publishers started to get interested in due to its deals with Google and LM’s such as Claude and OpenAI, YJN took its members directly to Reddit’s offices to ask how journalists should be using the platform, including how they should use it to source stories as well as ways to market themselves.

The Reddit event ran on a Wednesday afternoon, the kind of thing Duffy would have historically taken a half-day’s leave to attend. “Publishers were saying no, you take that as an offsite day; that’s really good training.” That, as much as any placement statistic, is her measure of success, one she concedes needs formalising. “That they understand what the difference between a diploma and a Master’s degree in journalism is, what a realistic entry-level salary in London is. Asking those questions that they maybe can’t ask parents or peers is how I measure success, not just purely on placements and jobs.”

The credential behind the network: communities over follower counts

Duffy’s day jobs are the reason publishers take her calls. At the Telegraph, where she rose to Deputy Head of Editorial Community and was nominated as a Rising Star at the AOP Digital Publishing Awards last year, her brief was reader retention: persuading journalists to spend twenty minutes below the line, mining the comment section for stories worth flagging to desks, getting reader questions answered on podcasts. Engaged readers, she argues, “become more loyal, and they also become advocates for your journalism. You’re not having to pay a marketing company or influencers to speak about your journalism. You have this amazing, powerful community of readers that are doing that for you.”

It was also where she built the Reddit expertise she discussed on a panel at newsrewired earlier this year. The Telegraph saw the platform as a route to a huge, largely American audience it could not reach on follower-driven networks dominated by the New York Times, the Washington Post, the Wall Street Journal, etc. On Reddit, she says, “it’s not about follower counts; it’s about communities”: the first account to post a major breaking story takes the traffic, whatever its karma. “In real terms, if done well, it’s a real win for referral traffic and page views. And I think that’s what a lot of publishers are coming around to now. Maybe a little late to the party, but good that they’re there.”

At The Times, the principle is the same at a larger scale. For this summer’s World Cup, her team (she credits Sifa Manara with the strategy) built a WhatsApp group of Times and Sunday Times readers travelling to the United States, Mexico and Canada. Those readers fed directly into the live blogs, sent pictures and voice memos when England and Scotland scored, and reported what it was like on the ground at the stadiums. Previously, she says, community journalism might have amounted to one article on Times readers going to the World Cup; this ensured that readers were at the heart of all editorial content and featured on The Game podcast.

The Media Stack is reader-supported. Join as a free member for occasional posts, or upgrade to £8.99/month for exclusive content and member benefits. Prefer to listen? Subscribe and stream via the Substack app.

The jobs that AI removes are the entry-level ones which they would have taken

The difficult backdrop is that many publishers are using AI to automate stories that would traditionally have been done by entry-level and trainee journalists. Bauer Media’s women’s weekly “Take a Break” recently stopped freelance commissions in favour of content written by AI; these are exactly the sort of jobs that typically would have been done by junior teams. Duffy does not dispute the threat.

“While AI is maybe removing some of those entry-level jobs, it’s also creating other roles that wouldn’t have existed previously,” she says, and wants AI literacy taught in schools, universities and training courses. Reddit, she notes, is “probably surfacing your work for you already” in LLM answers; understanding how is now part of your job.

She also highlights that the picture isn’t the same across the board. The Times aren’t removing any editorial entry-level jobs due to AI.

The access problem has not waited for AI. “I honestly think that anywhere more than an hour out of London is very, very difficult. It’s not even geography; it feels culturally very far away.” Entry-level roles have reached a gender-equal split, she notes, but senior roles are still predominantly dominated by men, and representation of BME (Black and minority ethnic) journalists, disabled journalists and regional voices is poor: “The numbers aren’t where they need to be or where they should be.” YJN points people to organisations such as the Journalists’ Charity and Women in Journalism, the former helping to fund travel to London interviews.

Bootstrapped, and choosy about the money

For its first 18 months, YJN was entirely self-funded, and Duffy turned down money. “We’ve had offers from a few companies, including some crypto ones, but it just doesn’t align with what we’re building.” The first significant partnership arrives on 25 August, when Qwoted hosts the network’s summer party at its offices. An agreement with Women in Journalism has also given YJN a regular newsletter slot and Duffy a place on their Committee. YJN will go international next month with its first event stateside, at Columbia Journalism School in September. The ambition is a YJN “wherever there’s a hub of young journalists that need support”, in the UK regions and beyond.

About 70% of her members want to be journalists in the traditional sense; 30% want the business side. Duffy’s own career suggests the line between the two is thinner than the job ads imply. “I didn’t know that my job existed five, ten years ago, but it’s so important to journalism. I still regard myself as a journalist; I curate stories and connect with audiences and people in new ways.

That is the double shift she has signed up for: keeping readers inside the industry’s paywalls by day, and holding its front door open on evenings and weekends.

The Cannes Lions Festival runs on the idea that advertising is still the main event. Rory Dolan, who is Head of Data Science and Marketing at Canva, is not so sure. He last came in 2019, roughly the last year the festival was really about advertising, by his reckoning, and the platforms have moved in since then. “The industry has a short memory,” he told me at Canva’s palatial beach pavilion on the Croisette. “Meta, Google, TikTok all turned up as partners. And then the next AI players are probably going to consume the industry again.”

Which is a striking thing to say from inside Canva. The company arrived at this year’s festival as a partner to exactly those companies. Its advertising product, Canva Grow, pushes creative straight into Meta and TikTok and reads the results back. Dolan knows the pattern he is describing. He is, after all, helping to build part of it.

What drew him, he says, was partly the mission. Canva sells itself as a company “fundamentally trying to do good”, pointing to the free products it gives schools and non-profits, while presenting itself as an antidote to visual sameness. Dolan describes an industry “reducing down to different versions of grey, white, black and brown”, a condition he calls chromophobia; Canva’s maximalist brand is his answer to it.

Rory Dolan, Head of Data Science and Marketing, Canva

The point of the rules is to stop reinventing them

Dolan started in options trading in London, moved into media agencies almost by accident, spent close to 15 years across planning, buying and strategy, built a consultancy within WPP’s GroupM (Now WPP Media), then crossed to the platform side and set up TikTok’s marketing science function across Asia Pacific. The common thread was measurement. His turning point was discovering the Ehrenberg-Bass tradition, the body of work that treats marketing as a discipline with enduring rules. “There are core rules. It’s not always changing,” he said. For a discipline he had heard dismissed as a soft science, that was a revelation.

At Canva’s scale, that belief leads to standardisation. Dolan runs a system across many markets, and his job is to ensure that what reliably works is not debated from scratch every time. “Let’s stop trying to reinvent the wheel on the things we know work and focus our time and resources on the fun part,” he said, meaning the ads and the brand.

Canva is opening its first internal marketing academy this year, bringing in experts to teach both the principles and the company’s way of working. Dolan also runs the behavioural science and capability teams, and he hires for skills the company does not need yet. He reaches for J.G. Ballard, of all people, for the idea that the future is already visible in the present if you are willing to look for it. Better to build the capability before the moment arrives than to go scrambling after it.

A global brand should be mistaken for a local one

Canva has moved away from single global campaigns towards work made market-by-market. In Dolan’s experience, global creative travels badly. “We’re trying to build things which are culturally and emotionally connecting to people, and it’s hard to do that building ads from Sydney, or from LA,” he said. The ambition is for the brand to look and behave like a local one everywhere it operates. In Brazil and the Philippines, he found consumers surprised that Canva was not a Brazilian or Filipino company, a measure of success that doubles as a boast.

Going local raises creative and resourcing costs, and Dolan concedes as much. He argues that the extra cost is justified by effectiveness, pointing to the messenger effect in behavioural economics: people trust messages more when they come from those who resemble them, and that resemblance goes beyond gender or ethnicity to social and demographic factors. The mid-market brands winning through local relevance aren’t, in his view, doing it to save money.

Humour keeps earning its place. The research case is that a strong emotional reaction builds memory; it also happens to suit a brand that wants to feel positive. Canva’s UK campaign this year installed Gemma Collins, the reality-TV personality, as a mock creative director, filmed in its London office, with an AI tone-of-voice toggle built into the product itself. Canva’s own European marketing lead has since described British audiences as more cynical and more modest than American ones, which makes much the same case for localisation. Dolan has a simpler test. “Would your campaigns come up at a dinner party? And would people laugh when they talk about it?” If not, the work isn’t good enough.

Brand sets the ceiling on growth

The harder sell inside any company is the finance team. Strip out the jargon, he argues, and brand is demand generation while performance is demand capture. “If we want to increase our rate of growth, we’ve got to increase the ceiling,” he said. Short-term measurement still matters to him; he cites unified lift studies that show a measurable effect of brand exposure on performance. Timing is harder. He invokes Mark Pritchard of Procter & Gamble, while acknowledging the wording may not be exact: today’s sales were built a year ago, so the job is to think about a year from now.

His preferred image for long-term brand building is an engine. “If you turn the engine off, the plane doesn’t fall out of the sky immediately; it glides,” he said. The catch is that a brand can coast down slowly, which is exactly why the decline is easy to ignore. Climbing back then costs far more energy than the glide ever saved, and that is the part he wants finance directors to understand.

The loop Canva is closing could close on itself

Millions of designs are made inside Canva every day, and the company is turning that into an advertising business. Canva Grow, relaunched at this year’s Cannes Lions, lets a marketer build creative, push it live and have the conversion data flow back automatically through platform integrations. Dolan is clear that the point is response data rather than pre-testing. “The real measure comes from when you put it in front of a customer and see what actually happens,” he said. Canva has also been buying its way further into this territory. The direction of travel is towards owning more of the path from creation to measurement.

Cheap, instant variation is one of the promises of AI creative, and it is also what makes Dolan wary. “Just because you can doesn’t mean you should build all those variations,” he said. He traces creative fatigue less to the ads themselves than to platform attention metrics that reward endless novelty. He has imposed rules to protect the quality of the signal, including a block on forced-view advertising, because the clearest signal is whether someone chooses to watch at all.

Dolan’s response is that Canva keeps the optimisation at arm’s length. The work of improving creative with the platforms, he says, sits “above the business”, built from the platforms’ own meta-analyses and shared playbooks “rather than us going through our customer data”. The aim is to start with best practice and let clients build their own expertise from there. He also points out that platform data can mislead in both directions: a weaker ad can outperform a stronger one “just because of how the algorithms work”.

A creation-to-performance loop risks rewarding whatever the platforms already reward. At enterprise scale, it will also need a measure of performance that stands apart from any single platform. Automate variation against platform feedback and you risk producing, at speed, the same kind of sameness Dolan blames on short-termism. This is the chromophobia Canva’s brand is supposed to cure, and the risk is that the loop he is helping to build could create more of it.

The people who translate are worth more than the people who know

Dolan gives his teams a particular instruction. He looks for depth in one or two areas on top of broad range and tells his people their job is to be “the second-best opinion on anything in the room.” Put those skills together, he reckons, and that makes someone one of the most valuable people in the room. The scarcer skill is translation: turning marketing metrics into language that finance, product and leadership teams understand. He hires for potential as much as completeness; if a role needs five capabilities, he will take someone with three and build the rest.

On AI he is deliberately understated. He expects next year’s conversation to be about products built for mass consumers rather than the industry talking to itself. “Anyone who knows what a context window is probably not part of the wider consumer audience,” he said. AI, as he frames it, should feel like magic and behave like a tool. Treating it as a category in itself misses the point, he argues, in the same way you would not call a hammer construction. Canva, he insists, doesn’t believe in replacing people. The point is to remove the barriers between what is in someone’s head and what they can make.

It is a comfortable position for a design company that frames itself as a force for good, and a harder one to hold while the same company closes the loop between creating an ad and measuring it, a process that has not historically required an agency to sit inside it. Dolan’s own reading of the festival was that every platform arrives as a partner and leaves having eaten something. Canva is a partner now. Whether the rules it is so carefully writing down protect creativity, or simply help industrialise its disappearance, is the question he has set himself. This year does not settle it.

The FIPP World Media Congress marks its 101st year this October in Madrid, and The Media Stack will be there as the Congress’s sole media partner. This is our first partnership with FIPP, and readers get something concrete out of it: a Congress pass booked with the code FIPPSTACK comes with two free nights at the Hotel RIU Plaza España, the Congress venue itself.

FIPP Congress runs from 13 to 15 October under the banner “FIPP 101: The Fundamentals of Media’s Future”, a programme built around what comes next, after a centenary year spent celebrating the first hundred. Organised with WAN-IFRA and Spain’s ARI, the Congress expects more than 500 delegates from over 40 countries. For three days, most of the international publishing industry is in one building.

Goli Sheikholeslami, CEO of POLITICO, and Laura Williamson, Editor-in-Chief of The Athletic, headline the programme. Others presenting include DPG Media’s Christian Van Thillo, Harvard Business Review’s Erika Heilman, TED’s Cyndi Stivers and DIE ZEIT’s Dr Rainer Esser, with more than 50 speakers confirmed and additions still to come.

Between them they run some of the most closely watched subscription, sports and B2B media businesses in the world, and the questions they will be answering are the ones The Media Stack exists to cover: AI licensing, the collapse of search referrals, what replaces the traffic economy.

The offer

Congress organisers have reserved 50 rooms at the Hotel RIU Plaza España for the Tuesday and Wednesday nights of Congress week. Book a pass using the code FIPPSTACK and both nights are free whether you are a FIPP or WAN-IFRA member or not.

The passes:

• FIPP/WAN-IFRA members: €1,500

• Non-members: €2,000

To put that in context: the standard member pass alone is currently €1,490, rising to €1,690 at final pricing. With this offer, roughly the same money buys the pass plus two nights in the building where the Congress is happening. No taxis, no 7am walk across Madrid, and the conversations that matter tend to happen in the venue bar after the sessions.

Fifty rooms will not last. Book at fippcongress.com with the code FIPPSTACK.

Please note our booking system does not allow for separate confirmations, so if you use the code will get a room confirmation within 3 business days directly from the hotel.

The Media Stack will be reporting from Madrid throughout Congress week: interviews, session coverage and analysis, here and on themediastack.co.uk. If you are attending and want to talk, get in touch.

Search historically offered a reasonable deal. You had a person searching, a results page popped up, a link was clicked and a site visited. Then the conversion, which may or may not have happened. The advertisers received the signals and while this measurement was never perfect, with look-back window differences or recency tossing a curveball, the actual journey had enough actionable data to support a decision. A business decision. Now this deal is breaking down.

People are no longer only entering “keyword” but asking full questions of LLMs. They’re adding context and following up on the answer, and the AI system does not retrieve a list of possible destinations but reads across them, synthesizes information and returns a response.

The search experience has changed from finding a source to getting an answer but the tech stack for measurement still relies on the same set of signals and assumptions, chief among them that value begins when a site is visited.

What happens when a simple query is no longer a reliable unit of intent

We can all agree the traditional query was imperfect but it was useful. “Best running shoes” could be classified, priced, compared and reported. Search teams could see the term, the ad, the click and the landing, and SEO teams could see an impression or a position and a visit. The query was a solid proxy for intent.

Conversational search is a different beast where a person may begin by asking for running shoes for heel pain, then refine by surface and brand, then reject a few recommendations based on price and finally ask which remaining options have the best return policy. This is not five isolated searches but one decision developing through a series of questions or a conversation.

But reporting does not see it this way. Google’s own Search Console documentation says that when a user asks a follow-up in AI Mode, it’s counted as a new query. This is consequential. The platform records another impression, click, position, etc. The person experiences continuity but the current measurement records separate events and the meaning sits in between the prompts.

Intent is being refined throughout the conversation while reporting continues to count its parts individually.

AI answers are creating influence without clicks or visits

The click has always been an incomplete measure of influence. We in PPC have argued for years that optimization of clicks or high CTRs is diametrically opposed to actual conversions. Now AI search is making this fact impossible to ignore.

Pew Research Center analyzed 68,879 Google searches conducted in March 2025 and the results when an AI summary appeared are astounding.

• People clicked a result link in 8% of visits compared with 15% when there was no summary.

• A link inside the AI summary was clicked in just 1% of visits.

• People ended their browsing session on 26% of pages with an AI summary, versus 16% on pages without one.

Those numbers are usually presented as a traffic problem for the publisher. And, make no mistake, they are. But they also describe an advertiser measurement problem. When an AI answer uses publisher reporting to explain a category or compare features of a product, the publisher may have influenced the decision without receiving the visit, and an advertiser may benefit from that recommendation without recording a referral, an assisted conversion or even an identifiable exposure.

The influence still happened; we just can’t confirm it. So publishers lose the visit, advertisers lose the attribution and the platform has interaction data that connects the two of them.

The available metrics describe visibility but not contribution

Here is where it gets a bit twisted. GSC can report clicks and impressions associated with links in AI Overviews or AI Mode, which is useful, but it preserves the old measurement framework:

• Was a link displayed?

• Did somebody click it?

It doesn’t show if a source helped shape an answer when its link was not visible. It can’t show which answer within the conversation changed the person’s criteria. And it doesn’t connect a cited comparison with a later branded search or direct visit or marketplace purchase … or, I’d argue, just as important, an offline conversion.

Even “impression” demands more scrutiny. Can a link record an impression just because it appeared within an AI response? This still doesn’t tell a publisher whether it influenced the response. On the other hand, content may contribute to an answer without producing an impression at all. So at the end of the day, visibility and influence are no longer interchangeable, if they ever were.

This creates dashboards that are “technically” correct but, for a business, utterly useless for strategy and decision making.

The reported clicks happened; the recorded impressions met the “definition” and conversions followed the rules but the decisions don’t conform to these rules and definitions.

More activity can conceal less understanding

The problem is not merely undercounting.

It is making the wrong business decision with data and, worse, confidence.

In one small paid-search account I reviewed, 255 clicks produced a 5.76% click-through rate at an average cost of $3.06. The account spent $780.79 and generated zero conversions. Okay, some of the engagement metrics suggest the ads were doing their jobs, but under the hood, you can see traffic drifting into searches for AI tools rather than the product or service being sold.

This example comes from paid search rather than an answer engine but it exposes the same lack of transparency. A click can prove that something was selected and at the same time, can’t prove that the platform understood the intent. The interaction had value, yes, but the data should absolutely not guide the optimizations, strategy or any budget decisions. It’s a house of cards.

As platforms automate query matching and AI takes over discovery, this gap between recorded event and intent grows. More queries, impressions and engagements are creating more data while giving advertisers and publishers little visibility into why a decision was made.

Publishers and advertisers now share the same blind spot

The old search gave publishers traffic in exchange for making their content available and that traffic aimed to support subscriptions or advertising. Now the AI systems weaken this value proposition by taking far more information than they give. Cloudflare estimated that in June 2025, OpenAI crawled publisher pages about 1,700 times for every referral it sent. Anthropic’s ratio was 73,000 to one. Ouch.

For publishers, this is audience and inventory that is no longer for sale; for advertisers, the loss is action data. If fewer people land on publisher pages, there are naturally fewer referral paths, assists/lift and audience signals used to evaluate the media. The same zero-click interaction removes value from both sides, even when the answers do their job and move people closer to buying.

This is why the issue can’t be left to the SEO, AEO (answer engine optimization), analytics and PPC teams independently because it’s not a channel-specific reporting problem.

It is a market-level loss of observable actions and engagements.

The market will continue to optimize towards what it can still see

Measurement systems do more than report performance. They also determine who or what receives the money.

If AI influence remains opaque, the budgets will continue to favor the last observable action: the branded search, the direct visit or the conversion event. The source that created the demand may look weaker and the channel may look more efficient even though that’s not the case. Publishers may inadvertently cut the content that influenced decisions because it can no longer produce enough traffic and advertisers may shift spending towards the touchpoints that remain measurable instead of the ones doing the most work.

This has become a strategic risk.

The industry is changing how people behave and discover and make decisions faster than it is changing how it values or even measures influence. We’re adding conversational interfaces, LLMs, generated answers and agents within a measurement architecture that doesn’t quite fit because the buying journey or “path to conversion” is no longer linear.

The result will not be the absence of measurement. It may be worse.

Precise measurement of a shrinking portion of the decision.

This is a guest post written by Shelley Stone, follow her Substack at

Related Articles

The UK economy grew 0.6% in the first quarter of 2026. UK advertising grew 9.3%. The latest Advertising Association/WARC Expenditure Report, published on 30 July, puts total investment at £11.7bn for the quarter, the strongest growth since Q4 2024 and 1.7 percentage points ahead of WARC’s own April forecast. After inflation, the rise is still 6.0%. The full-year forecast has been raised to £50.5bn, an 8.2% increase on 2025.

A market growing many times faster than its economy demands an explanation, so The Media Stack put the question to the people who compile the numbers: Stephen Woodford, Chief Executive of the Advertising Association, Suzy Young, Partner Relations at WARC Media, and Alex Brownsell, WARC’s Head of Content. The answer is a structural shift in who advertises at all.

The growth is coming from three million businesses that didn’t advertise before

Woodford’s starting point is that the disconnect is real but overstated. “The overall economy is sluggish, but advertising is clearly really dynamic in terms of its growth,” he said. “We are a very digitally advanced economy, in large part because we’re a very advanced e-commerce economy. The UK is number three in the world in terms of investment in AI and by far the European lead. I think there is a very British tendency to talk ourselves down.”

His fuller answer takes a 25-year view. In 2000, on his numbers, the UK had about three and a half million private sector businesses, and around 300,000 of them advertised. “Call it 8% of the businesses in the UK.” According to their Advertising Pays 2025 report, there are now around five and a half million businesses, and the number advertising has grown more than tenfold. “So around 60% of the businesses in the UK now use advertising.”

“Whether you’re a local builder or pet shop, or a giant multinational, and every type of business in between, you’re going to be using advertising to reach your customers,” Woodford said. “That advertiser growth has been another factor driving the growth in the ad market way ahead of the growth in the economy.”

That long tail advertises almost entirely on the platforms; some 90% of small businesses use digital advertising, by Woodford’s count. It explains the shape of the quarter: search up 9.8% to £4.6bn, social media up 17.7% to £3.1bn, retail media up 17.9% to £891m. The performance channels took roughly £8.5bn of the £11.7bn total.

Even the traditional channels are growing digitally

The report’s other pattern is that growth inside legacy media is digital growth. Digital radio was the fastest-growing line in the entire dataset at 22.1%; digital out of home rose 17.6%; addressable TV climbed 15.5% while total TV managed 0.8%. “Even in the traditional media sectors, it’s the digital formats that are growing fastest,” Woodford said. “The ability to use data and segment and become ever more efficient in your targeting is a key driver, which is absolutely not to say that broad reach media doesn’t still have a really important role.”

He is also wary of treating the headline number as a single market. “Historically, we’ve looked at broadly a £50 billion total ad investment as if it’s one homogenous market. Actually, it’s a hugely diverse market.” That concern will produce something new next April, when the full-year 2026 figures are published: “We’ll look at the spend pattern of the very large advertisers, classically the ISBA members, the advertisers that use IPA agencies, versus that long tail, but also try and segment the long tail between the micro businesses and the medium-sized businesses.”

A survey running since 1982, argued over every quarter

The Expenditure Report is cited from conference stages as the industry’s gold standard, but few of the people quoting it could say how it is made. Young, who compiles it, was pleased to be asked. “It stems from media owner surveys. It’s something the AA and WARC started all the way back in 1982, and the numbers have been compiled quarterly since then,” she said. WARC surveys the publishers directly, with digital revenues incorporated since 2011, and works with the trade bodies for the rest: Thinkbox, Radiocentre, Outsmart, IAB UK, Royal Mail and the Cinema Advertising Association. Forecasts fold in economic outlook, forward guidance, agency data and short-term factors such as elections and World Cups. This summer’s tournament is expected to help push Q2 growth to 8.9%.

Incoming data is sense-checked against the previous quarter before anything else. “Quite often we have methodological changes, so we have to be really careful that we’re ensuring the data are like for like,” Young said. Definitional disputes, such as whether YouTube watched on a television set counts as TV or digital, go to a quarterly forum where all the trade bodies argue it out. “It’s got to be agreed as fair by everybody.”

The categories themselves are still moving. Retail media and social media were broken out publicly for the first time in April. A pure-play online audio line, covering podcasts, is being worked on by the IAB and Radiocentre. And WARC is talking to the influencer marketing trade body about a line for the creator economy, with no illusions about the difficulty. “There’s no unified billing system. There’s no unified metrics,” Young said. Woodford is confident a number will come, “but it will be a process of gradual refinement, because it is such a complex market that doesn’t necessarily fit neatly into conventional definitions, with a blend of media, production and talent spend.”

Retail media has not peaked, but outside Amazon it is already slowing

Retail media grew 17.9% in the quarter and WARC forecasts 15.9% for the year. “I certainly don’t think it’s peaked. There’s a lot of headroom for growth,” Brownsell said, with the US market showing where the UK is heading. The qualification concerns who captures it. “Because of the dominance of players like Amazon, global retail media growth is now dropping down into single digits for the first time away from Amazon.” The task for Tesco, Sainsbury’s, Boots and the other local operators, all of which he sees doing “really exciting things”, is to capture share of budget against it.

Trusted news brands sat out a digital boom

The one corner of the market still shrinking through the boom is published media, down 5.9% in the quarter. National newsbrands fell 3.7%, and their digital revenues fell too: down 0.8% in a quarter when every platform channel grew by double digits. For the most trusted brands in British media, that is a bleak scoreboard.

Brownsell does not read it as a verdict on the journalism. “They have a challenge to tick certain boxes that the big tech platforms have been able to do, in terms of measurement, ability to target audiences, the optimisation side of things. But what they do have is that premium environment. They have the trusted audience engagement.” And the global picture is turning: “Having seen a precipitous decline over the last 10 to 15 years, we are now seeing stabilisation in ad spend with news brands and magazines, and actually we’re forecasting growth in global news brand ad spend next year.”

The longer-term threat is the one mapped in WARC and PHD’s From Abundance to Agents, which projects media and publishing as one of the categories most exposed to agent-mediated commerce. “Consumers in general seem to trust the algorithm to find the content they are likely to find most entertaining,” Brownsell said. His advice splits by scale. “Where there are strong publishing and media brands, it’s about doubling down on those direct relationships and, where possible, preventing disintermediation. For the longer tail, it is going to be a case of learning the new rules of distribution.”

Woodford sees a countervailing force. “One of the things the LLMs use to make recommendations is authoritative editorial content. What the Economist says will rank very highly. What third parties say about you becomes increasingly important.” In a world of machine-made abundance, he argues, “trusted brands, trusted voices just become even more important than they are now.” The economics, he concedes, lag the technology: “As technologies emerge, the economic models have got to catch up with them. There is a symbiotic relationship between them. If one in effect undermines or destroys the other, it ultimately undermines their own business.”

A Burnham bounce, in the report’s own words

The report itself wonders whether there will be “a noticeable ‘Burnham bounce’” in the second half. Woodford, asked about the new Prime Minister, reached first for stability: “Regardless of your politics, we need him and his administration to succeed. There are signs of a Burnham bounce in consumer confidence, which is a good thing to see.”

The optimism rests on Burnham’s record. A former Culture Secretary, he made the creative industries one of five priority sectors in the North West, where the MBacc (the Manchester Baccalaureate, Greater Manchester’s technical-education route) paired industry with education authorities and universities. “Manchester is the second city for advertising after London,” Woodford noted. James Purnell, another former Culture Secretary and a BBC alumnus, arrives as Chief of Staff, and Lisa Nandy stays at culture, which is the continuity Woodford wants. “We don’t want to see an ever-changing cast of ministers in key roles.”

The forecast has one standing risk attached: the intermittent Middle East ceasefire, and the chance that a protracted conflict feeds back into energy prices and inflation.

For now, the strongest quarter since the end of 2024 rests on millions of new advertisers buying performance media from a handful of platforms, while the country’s most trusted media brands wait for the machines that already read them to start paying them. The economic models, on Woodford’s telling, always catch up eventually.

Related articles

Jon Morgenstern’s job title is EVP, Head of Investment, which sounds more like a venture fund than a media agency. He knows it. “I do get confused all the time for being Gary Vaynerchuk’s VC person, for if he wants to invest in things,” he says. “I disappoint a lot of people regularly.” The reality is simpler: “Investment, in the way we use it, is a fancy term for media buying,” stretched in practice across media investment, technology, partnerships and measurement, alongside VaynerX’s own infrastructure.

The title is a small version of a much bigger problem. Organic and paid, working and non-working, influencer and creator, buyer and engineer: these are the words the industry uses to plan, price and pay for media, and the platforms are steadily emptying them out. They are also written into measurement frameworks, procurement contracts and job titles, so an industry that broadly agrees they no longer mean much still runs its business on them. Morgenstern has spent eleven years inside VaynerMedia watching it happen, and has given up waiting for anyone to supply replacements.

What a $3 CPM actually buys

The pressure lands first on measurement. A social-first agency asking for a bigger share of a budget gets one question back from the finance director: prove it. The industry’s own vocabulary makes that hard to answer. “YouTube can show up in a social line item, a TV line item, video, online video, working media, non-working media,” he says. Influencers count as working media at some brands and non-working at others, depending on how the fees are wrapped.

There is a crude way to put a number on organic: divide what the content cost to make by the impressions it earned, and call the result an effective CPM. Morgenstern rejects it outright. The method “very much under-represents the value. It assumes that an organic impression and a paid impression are equally valuable.” They are not. Cheap reach, bought at “bargain-basement CPMs, like a $3 CPM”, is inventory nobody else bid for. An organic impression is one the platform chose to serve, on the grounds that “out of everything else on earth we could serve to this user, it should be this highly relevant piece of content”.

That judgement is what makes his bigger claim work: organic performance is now the most honest research tool a brand can buy, because on this one point the platforms want what the advertiser wants. “The only thing the Metas and TikToks care about more than ad revenue is keeping people in their app, paying attention and scrolling.” Content that “passes the organic great filter” on TikTok, Reels or Shorts, where traction no longer requires followers, has survived a test no research panel can reproduce. The emphasis is on reallocating research budgets rather than abandoning research altogether: “Stop paying for as many focus groups and storyboards and all of these things that go into legacy media.”

That conclusion happens to suit an agency built on organic-first creative, and it comes at a price: if the organic filter sets the standard, the standard is being set inside Meta and TikTok, on ranking systems built to keep people scrolling rather than to grow anybody’s brand. He is less willing to apply the same doubt to the attention data he wants finance directors to treat as proof, which is the open question under an otherwise convincing case.

The contract still says 80:20

Suppose the finance director is persuaded. The contract still is not. Client appetite for change is the highest Morgenstern has seen, with the reference point now set by brands born on social: “We’ve kind of squeezed as much of the juice out of the legacy model as we can.” The blocker is procurement. Agreements still fix working to non-working media at ratios like 80:20, and creators, content production and AI tools, the things clients now want more of, sit on the wrong side of that line. “We’re not saying go to 50:50 just to be wasteful. But it requires a lot of education and consensus building. It has been getting done, honestly, because the pressure is on and folks know it intuitively.”

AI has sharpened the same standoff and flipped who is asking whom for a discount. “A lot of how it manifests itself to us is negotiating with procurement, and they’re like: we want you to be 15% cheaper, and still give us the same or better output, because AI.” He is clear about what it does to an agency that moves fast to satisfy it: “Any agency that doesn’t feel very cautious or hesitant to unleash an AI tool, they should feel hesitant.”

The negotiation may also be beside the point, because procurement rarely decides which tools get used: “If you use Microsoft Teams and Microsoft Suite, it’s going to be Copilot. We use Google Workspace, so it’s going to be Gemini Enterprise.”

Gary opens at a ten and expects to land at four

Internally, the loudest voice is the one not to take most literally. Gary Vaynerchuk’s line has long been that organic views are the ultimate measure of creative success and that television can be a waste. Morgenstern laughs at how that lands on a client call: “Gary will come in and be like: zero programmatic, no banner ads, no display. And then the media folks within Vayner are on the call with the day-to-day clients after.” He squares it with a formula: “He says the industry’s at a one, I come in at a 10, expecting it to land at a four.”

Opening high is a negotiating tactic, used on his own agency as much as on the market; you can only rip up so much of a media mix in a year.

Read closely, it is also an argument about price rather than about channels. “He’d buy any media channel or tactic, including TV, at the right price. It’s just a question of overpriced or underpriced relative to the actual demand creation. He always says these are all potential reach figures, not actual reach.” Which returns the argument to where it started: a $3 CPM is only cheap if the reach behind it is real.

Subscribe now

The follower moat has drained

If the categories no longer hold, the transactions underneath them have to be renamed. Vayner’s creator practice is called CAIT: creators, affiliates, influencers and talent, four words for four different deals. “Influencers have media distribution. You’re paying for them to make the post and post it on their handle, to their people.” Creators are hired for the content itself, which the brand can run as paid media and measure against acquisition costs. The influencer end is changing fastest. “We’re in the non-follower era. The moat of just having a huge following, and whatever you do having that baked in. The cult of personality still matters, but share-wise, more is going to creators and talent.”

Take distribution out of the deal and what is left to buy is raw material. Vayner uses tools to find local creators near individual restaurant franchises who are “not even on the hook to film it, edit it, make it into a Reel or a TikTok, post it. Just give us the footage, and we’ll pay you for the footage.” Its own “views editors”, part of the clipping industry that has grown up around short-form video, do the rest. It removes what he calls “so much of the dead weight loss in the creator economy”.

The company keeps rebuilding around a dead category

This is not the first time the company has rebuilt itself around a category that stopped working. Morgenstern arrived in 2015 from SocialCode, when “paid social was trivial” and his biggest clients were agencies outsourcing it. What drew him to Vayner, then a creative and organic social shop, was the finger-pointing everywhere else: “we’re pulling every lever we can, the creative is the bottleneck. The creative agency would be like, well, we delivered beautiful creative. Your media buying shop must be screwing up. No accountability.” That grievance became the pitch. “When we have both, we are accountable. There’s no one to point to but us.” Integration is unremarkable now; in 2015 it was a way of taking away everybody’s alibi.

The second rebuild finished this year, in what he calls “this big growing up of sorts”. In January, The Sasha Group became ChukMedia, “now really a proper conflict shop”; in April, VaynerX added Tamara Group, built largely from Gallery Media Group’s publishing operation. The driver was conflict, and the separation cost more than expected: “we might want two clients in this certain category, but really need to be airtight and do the whole actual separation, which is more involved than I certainly realised.” An agency that can hold two rival accounts in the same category is being paid for something other than a channel remit, which is the 2015 bet made bigger.

The media buyer becomes a systems plumber

The last category to dissolve is the one on Morgenstern’s own trade. A decade ago, serious dynamic creative work needed an engineer or an ad tech vendor. Now, he says, the future media buyer swaps manual spreadsheet work for code:

“More of an engineer, or a systems plumber. True deterministic: this report will be this way every time. There’s no ‘LLM, what happened here?’”

The rule inside Vayner is “human upfront, human at the end”: the briefing, strategy and context go in, a person checks the work and presses go, and the boring middle gets automated first.

He has been here before. In 2011, before newsfeed ads existed, Facebook advertising had no playbook. “We were writing the rules and figuring it out with clients, and it was fun. That’s kind of where we’re at today with AI. … Everyone’s waiting for someone else to say go.” His advice is blunt:

Every category Morgenstern has watched dissolve since 2015 took a manual down with it, and each time the industry’s instinct has been to wait for a replacement from the same platforms that removed the old one. Measurement, contracts and job descriptions are all still written in the old words. “Had we not evolved, we’d be in trouble,” he says. “Standing still is definitely death, more than ever.”

“You will read more about Australia than you will about Indonesia,” Duncan Hooper says. “Indonesia is the fourth largest country in the world... It’s virtually a major power.” Yet in the British press it surfaces only when there are floods, or a story involves holidays in Bali. “We’re still in this world where you are expecting news from these countries that you haven’t heard of to either be a war or a coup or a natural disaster.”

​That blind spot is the business case for Global South World, the digital publisher Hooper serves as Consulting Editor. Founded in 2023, it covers the emerging economies of Africa, Asia and Latin America (for readers in those countries as much as for the West) on the theory that media’s energy has moved to places with young populations and growing markets while Europe’s industry contracts and turns inwards. “I don’t want to say it’s moribund. That’s not fair, but it’s facing major challenges,” he says of the business he came up in.

Hooper’s route to that conviction runs through most of British media. His first job was at J.P. Morgan, “and I hated it,” he says. “I noticed there were these things called Bloomberg terminals, and I saw this news on these things, and I thought, wow, I bet not many people who are graduating in English know that Bloomberg terminals exist and that this is a company that does news. I applied to all the normal places and got rejected from all the normal places. And I wrote to Bloomberg, and they gave me a job.”

From the newswire he moved through Euromoney magazine, the Telegraph’s digital operation, MSN, Euronews (where he set up Africanews, the first bilingual pan-African television channel) and CGTN, the international arm of China’s state broadcaster. That last posting shaped the thesis behind Global South World. “One thing about China is it’s very interested in the whole world,” he says. Working alongside colleagues from across Asia, Africa and Latin America showed him development that Western newsrooms barely register: “The numbers are just frankly astonishing in terms of the proportions of people who are firstly being brought out of poverty, but then also being brought up into the middle class... And they need information. They need reliable, credible information.”

​The coverage that does exist skews grim, and Hooper thinks the gap between perception and fact has widened. In a recent quiz, he asked what has happened to forest cover in China over 20 years; it has roughly doubled. Costa Rica has rebuilt decimated rainforest into a national asset. “There are a lot of positive stories coming out of these countries,” he says. “And I don’t think those stories get heard enough.”

The only person in the UK is the consulting editor

Global South World sits within the media division of Impactum, alongside the French analysis site Atlantico (fifteen years old this year, and subscription-funded), the social-first brands The World in Maps and World Visualised, and Mondance, a recently acquired dance sport app. Hooper describes his own role as “a troubleshooter”: strategy, product, hiring, training, analytics, “wherever I’m needed.”

He is also, deliberately, the only person in Britain. “None of the other staff are based in the UK. It was a very conscious decision that we wanted to work with people from the countries that they were reporting on.” Editor Edward Sakyi runs the site from Accra with a team of around six; a correspondent in Manila covers Asian hours; a colleague in La Paz does the same for Latin America. Around that core sits a looser network of news content creators on LinkedIn, mainly ex-journalists, providing perspective from their own countries.

Distribution follows the audience rather than the desktop homepage. The group’s younger brands are vertical, mobile-first sites designed for minimal friction, because Hooper believes the subscription model that sustains Atlantico is out of reach for a new brand. “If you have a younger media, it’s very, very hard to grow on that model. So we feel that you need to grow through social media.” A Portuguese-language edition launched in Brazil in early July, built with Giraffa Tech, the group’s media-tech partner, with further translated editions in testing.

The audience skews educated: people in international business, or those whose English marks them out. The US tops the analytics, how much of that is VPN traffic is hard to say, with big audiences in Indonesia, Nigeria and South Africa. The appeal, Hooper argues, is the outside perspective itself. When the site covers Ethiopia, Ethiopians read it. “They like to see how the rest of the world is talking about us, and they don’t see it enough.”

Subscribe now

This disruption will be bloodier because nobody has reserves

Hooper lived through the first digital disruption at the Telegraph. “We went after page views. We wanted to be the biggest.” He thinks the AI wave will cut deeper. Not worse, he insists, but “more transformative, more disruptive,” for a structural reason: last time, incumbents had fat to burn. Media companies “were much better able to respond because they had reserves... to do some experimentation, to take their time a little bit to adapt, to take some hits, to get it wrong.” Few now have that cushion.

His own newsroom has no such cushion. “We don’t have an AI specialist who’s going out on the frontier. We’re not building our own models. At the same time, the off-the-shelf models are fantastic for us.” He has Claude pulling YouTube statistics weekly and analysing which categories and video lengths perform; AI generates ideas from past performance, assists photo montage work, and error-checks every article before publication, a discipline introduced after writers covering unfamiliar countries made “silly mistakes”. Abstaining was never an option: “We would just not be able to run our business effectively.”

The threat is just as concrete: competition from fully automated content, and advertising inventory that becomes “infinite, basically”. What worries him most is the licensing economy forming above his head. AI content deals are being struck with the biggest publishers, and it is easier for an AI company to sign Le Monde and the Guardian than to do “500 deals with smaller publishers”. He wants collective mechanisms (he cites the licensing standard RSL) to make sure the smaller layer of publishers, “who bring the diversity... who bring the plurality that we all agree that we want in media”, are not left without a seat at the table.

Plurality, not impartiality

Much of Global South World’s patch is served by media owned by mining companies or political interests. Ask Hooper how an impartial outsider reports on that, and he declines the premise. “I think everybody comes with certain biases in their reporting,” he says, noting that Britain has its own political and advertiser influence over media, merely with stronger regulation and, crucially, more plurality. Nor does he think claiming neutrality helps: “Organisations that we would say are being impartial over here in the UK are not seen as being impartial in other parts of the world.” Western coverage of Ukraine, he points out, looked anything but neutral from the global south. “It does endanger your credibility if you try to claim that you’re able to do that.”

His remedy is disclosure and competition rather than prohibition. A mining company owning a news organisation is not, for him, the scandal: “I think it should be explicit that the media organisation is owned by the mining company. People should be able to make their own decisions about it.” The real danger, he argues, is big tech deciding which outlets get paid and which get seen, killing off the blogs and niche sites the open internet allowed.

​TikTok is his case in point: his biggest platform, and the one that refuses his money. “You want me to rank the social media sites by how bad they are?” he says, when asked whether it is worse than Facebook. Facebook pays a little; TikTok pays nothing, excluding the brand from its creator programme because news footage of world events is deemed “unoriginal”. Its moderation systems flagged his own interview with an Indonesian minister as unoriginal, blocked a genuine video as AI-generated, and in Brazil blocked the company’s paid advertising over a light-hearted graphic elsewhere on one of its sites. “I’m not asking them to pay for it, even. I just want them to allow us to pay to give their audience the news.”

The business plan is deliberately slow: emails collected but no newsletter yet, an ad-funded model, a low-cost base and self-funding within two years. By 2029, Hooper wants Global South World to be a recognised name in Brazil, South Africa, Nigeria, Ethiopia, Egypt, Indonesia and the Philippines: on its site, its podcast, WhatsApp, YouTube, “even if they’re reading us on an AI platform... We’re platform neutral.”

​

First, the plainer target: “I want to be still around.”

​

​

​

A year ago, at Cannes Lions, Matty Beckerman was showing pilots. This June, aboard the IRCODE yacht in the Old Port in Cannes, he was scanning a live television broadcast from Salt Lake City and pulling up the attribution data behind every advert in real time. In between, Sinclair, the second-largest station owner in the United States, wrote him a cheque.

Beckerman is the Founder and Chief Executive of IRCODE, a computer vision company built on a simple premise: the image itself is the code. Point a phone at a television screen, a billboard or an album cover and the technology recognises what it sees, no QR code required.

The underlying patents were filed in 2015 by Philipp Holzschneider, a Berlin-based engineer with a background in drone vision systems who now serves as the company’s Chief Technology Officer. Beckerman acquired the intellectual property around three years ago and has been building on it since.

“We’re creating a digital layer to the physical world,” he told me. “Whether it’s in real life or watching a television show, you scan with the lens and get all the information behind it.”

When we first spoke in 2025, the obvious question was defensive: how long before Google, Meta or Amazon simply absorbed the idea? Beckerman’s answer this year was to point at the filing cabinet. “We are filing new patents almost once a month,” he said. “The nice news is our patents have been around since 2015, are in 25 countries around the world and are fully issued. They’re not pending.”

It is a lawyer’s answer to a product question. Apple and Google have both shipped visual search of their own since those first filings, and whether ten-year-old patents constrain products that already exist is a harder test than whether they deter copycats.

Will.I.Am explains how IRCODE works, on the IRCODE Yacht at the Cannes Lions Festival.

Salt Lake City does what pitch decks cannot

Beckerman’s evidence was a live feed. He brought up KSL, Bonneville International’s NBC affiliate in Salt Lake City, and scanned the broadcast as it aired.

“The entire broadcast for KSL is now scannable and interactive, the whole thing start to finish,” he said. “Rather than talking about deployments, this is actually in the market and working, and we have users doing this every day.”

Matty Beckerman, Founder & CEO IRCODE

The KSL Lens sits inside the station’s existing news app. Viewers point their phone at the screen during a bulletin or an ad break, and the app recognises the content, surfacing programme information, offers and competitions. The station promotes the feature on air, including a weekly $1,000 prize draw for users who scan commercials.

My own research ahead of the interview put KSL at around 18,500 scans with a 33% average click-through rate. Beckerman did not dispute the figures. “It’s so far beyond a QR code scan that it’s kind of astounding,” he said. “We’re seeing a 30%-plus click-through rate, which is a huge deal for us.”

These are strong numbers from a small base. Salt Lake City is a market of roughly a million TV households, and 18,500 scans, generated with heavy on-air promotion and a cash incentive behind them, is an encouraging start rather than evidence of a habit.

Nobody yet knows whether the scanning outlasts the novelty and the prize money, and finding out will cost a

station group real money.

Subscribe now

Sinclair’s investment turns a demo into infrastructure

Sinclair has decided to spend it. Two days before we spoke, the group announced a strategic investment in IRCODE, undisclosed in size, alongside a commitment to roll out IRCODE Lens across its stations.

Sinclair operates the second-largest television station portfolio in the United States, 185 stations across 86 markets, reaching roughly 40% of American households.

“We’re starting with two stations in July, in Austin and Salt Lake City,” Beckerman said. “Then we have ten more right behind that, and the plan is to be able to do this across their entire ecosystem.”

Local station sales teams have spent a decade losing advertising budgets to digital platforms that can prove outcomes. Scannable broadcast gives them something to sell back.

“The ad sales team can now go out and speak to all their local advertisers and say, we can give you real attribution, we can reach your consumer directly,” Beckerman said. Because the interaction happens inside the station’s own app, the engagement and conversion data stay first-party, owned by the station and its advertisers rather than a platform intermediary.

He puts the wider ambition plainly. “I would like to be on every television set, at least in the United States, by next year, and then expanding to the rest of the world.”

The consumer app is dead; the camera is the prize

A year ago, IRCODE was preparing to launch a direct-to-consumer app. That plan is gone. Beckerman explained the switch from chasing consumers directly to selling the technology to other businesses.

“We were going to focus on B2C, building our own consumer app and getting it out there in a major way,” Beckerman said. “But we found that white labelling our technology into existing apps is the best way to go. That’s the way people want it. You’re not having to download a new app.”

Amazon has Amazon Lens, and Google has Google Lens; IRCODE now builds lenses for everyone else, branded as theirs, running inside apps their audiences already have. That positions IRCODE as infrastructure and explains why the pipeline runs through TV networks rather than agencies or brands.

The ambition runs further, into the smartphone camera itself, which is where QR code reading eventually ended up.

“Pre-COVID, you had to download a third-party app, and then it went into the camera,” Beckerman said. “We’re hoping that’s our path.”

That path runs through Cupertino, and Beckerman has been laying groundwork. IRCODE has built for the Apple Vision Pro, and he says Apple has granted the company camera access on the device, something Apple rarely extends. He has also built for Meta’s Ray-Ban glasses and Snapchat’s Spectacles.

The attribution claims survive questioning, mostly

IRCODE’s pitch leans on deterministic attribution, a phrase the industry has learnt to distrust. Beckerman’s response was to show the back end.

In front of me, he pulled up the KSL analytics, individual scans registering in Utah in real time, then drilled into a single local advertiser: every airing of their spot, every scan against it, every click that followed.

The data is anonymised for privacy compliance, he said, but it is structured at the user level. “It’s no longer theoretical. This is a user doing this. Finally, this is real attribution data.”

The company’s marketing materials make bolder claims, including uplift figures of five to ten times for gamified and host-endorsed content. I asked whether those numbers came from controlled tests or live markets.

Brittany Hershkowitz, IRCODE’s Chief Marketing and Brand Officer, gave a more careful answer: “That’s a combined gamification number. One was in a controlled environment; one was in a live environment.” Buyers should read the headline multiples accordingly.

What the system cannot yet see is the whole funnel. Networks can measure engagement up to a point, after which the brand owns the data, and neither side can see across the join.

Hershkowitz said the company is building towards closing that gap, though she would not discuss details. “You’d be able to see the entire funnel, all the way from the moment you scan to an abandoned cart.”

If IRCODE delivers that, the claim stops being about interactive TV and starts being about measurement infrastructure. That is a far bigger market, and a much harder one to win.

Interactive TV stopped needing the smart TV

IRCODE’s lens works against an ordinary over-the-air feed, which is where the threat to connected TV lies. That market is where most of television’s advertising money is now heading. “You don’t have to have a connected television,” Beckerman said. “Now we make everything connected.”

CTV’s commercial pitch rests on the operating system in the set: interactivity, addressability and measurement as features of the platform. A phone camera that makes a 15-year-old panel shoppable moves those capabilities from the television to the viewer’s pocket, and hands the resulting data to the broadcaster, not the device maker.

For the companies that sell smart TV operating systems as the gateway to interactive advertising, that is an awkward proposition.

The QR code is the named casualty, but the platforms are the real fight

Asked who loses as IRCODE scales, Beckerman did not hesitate. “I’ll tell you who does lose in this: anyone who wants to keep a QR code on the screen. I think that’s going to go away.”

Craig Kallman, who ran Atlantic Records for three decades and is now Chief Music Officer at Warner Music Group, invested in IRCODE around six months ago. Beckerman relayed Kallman’s verdict with obvious relish: “He said, man, I feel like it’s the hair metal days and you just played me Nirvana’s demo, and no one realises what’s going to happen.”

The KSL numbers are promising, if small. The Sinclair rollout is barely a month old. And the big platforms have not stood still since Holzschneider filed his first patent.

Amazon and Google operate lenses at a scale IRCODE cannot match, and Apple now ships Visual Intelligence natively on recent iPhones, a long press on a side button that searches whatever the camera sees. Beckerman wants IRCODE to end up inside that camera. Apple has already filled it with its own.

Related Article

Marc Fischli opens with what Criteo is not. “People still think we’re a re-targeter, which obviously is the birth of Criteo,” he says when we sit down in the Criteo yacht at the Cannes Lions Festival “You arguably could say that we were one of the inventors of the retargeting piece.”

That business made Criteo’s name and then nearly dated it. When Megan Clarken arrived as Chief Executive some seven years ago, the third-party cookie was already under threat, and she pivoted the company towards retail media, largely through acquisition. That business remains one of its two pillars. The second rebuild came about 18 months ago, when Google finally decided against killing the cookie. “That freed up a significant number of engineers, over 100 engineers, pretty much overnight,” Fischli says. The result was a strategic shift back into performance media as a growth engine, coinciding with the arrival of current Chief Executive Michael Komasinski. The products now reaching the market are what that cohort of engineers has been building since.

Fischli runs Criteo’s international markets, covering EMEA and Asia Pacific with a focus on performance media. He joined three and a half years ago after a long spell at dunnhumby, latterly as Chief Client Officer, and Diageo before that, a mix he says now runs through Criteo’s senior leadership: martech specialists alongside retail, CPG and agency backgrounds.

“The battle for the direct contact with the consumer, or the traveller, or the user, is on,” he says. “At the beginning, the LLMs wanted to go into that. I think they’ve taken a bit of a step back, because the retailers are, as they should be, fiercely defensive around that. But the battle is on. Where that transaction happens is where the best data is. Who owns that, and how exclusive it is, is going to be quite important. If everyone knows everything, the value of that goes.”

Strip the company back, and Fischli sees two assets. The first is its commerce graph, which he says knows what people do on the open web and across apps well because everything gets tagged. The second is access to the product catalogues of its retailers and e-tailers at SKU level, alongside 235 retailers on the retail media side, some of which give Criteo visibility into what happens offline in their stores.

“That is probably something we have underutilised as a company for a while, but we are not a data sales house,” he says. “It’s not our data at the end of the day. But what we can do is use this data to have advantaged services for our clients.” In practice that means targeting, creative and, increasingly, measurement and incrementality.

The product architecture built on those assets runs in three directions on the performance side. Criteo GO is a self-serve, cross-channel platform aimed at small and mid-sized advertisers and at the tail brands of large ones. “If you’re a Unilever, you have 3,000 brands. The last 1,000 never really got a proper outlet,” Fischli says. An enterprise version, co-built with agency partners and in beta, is due in 2027, with curation through Criteo’s own and third-party SSPs as the third leg. On retail media, the company is moving everything to auction-based buying and positions Commerce Max, its multi-retailer buying interface, as its answer to fragmentation.

Everyone talks nice at conferences. Then Monday morning happens

At a panel earlier in the week, Fischli was asked why Europe is ahead of the US in retail media. “The answer to that is, of course, in-store,” he says. The American industry started with sponsored products and stayed focused on them. In-store is messy, and the budgets sit with retail commercial teams inside yearly joint business planning negotiations rather than with marketing. The Europeans started in-store. “They’re used to messy stuff. When brands wanted to activate all of that universe, from sponsored product to a screen that happens to be in an aisle somewhere, the Europeans were actually quite well prepared.”

Both sides share a deeper structural problem. Retail media inverts a relationship retailers have spent decades enforcing. “In a retail brand relationship, the retailer normally calls the shots,” Fischli says. “Now retail media is different, because the client is inverse. The brand’s the client of the retailer, and that is a mindset change that people still need to go through.” Walmart understood it early with Walmart Connect. Many retailers in the US and Europe have built retail media networks that sit in marketing, “and it’s not particularly powerful.”

Advertisers have their own version of the gap between intention and behaviour. “The true ‘I get fantastic value out of retail media’ is still more the exception rather than the rule,” he says. “Everyone knows it, and everyone sits at the conferences and talks nice. And then we go back to our offices, and everything goes back to normal come Monday morning.”

Strategy, in his diagnosis, is not the problem; boards are on side, and CFOs love the EBIT impact in thin-margin grocery. Structure and people are. “What is in it for me if I run the commercial team? As long as I do my number, I don’t need to worry about the other guy.” On his maths, if retail media captured even half the share of voice it should, the industry would be two or three times its current size.

Subscribe now

Partnerships and the honesty problem

Criteo’s retail media pillar was built by acquisition at speed, an urgency Fischli says no longer applies. M&A will continue, but selectively and with a higher hurdle. What has changed is the default. “What is the absolute new normal is partnerships. Everyone in the industry has realised, maybe with the exception of the very large American tech companies, that they cannot do everything themselves. You need to be much clearer about what you’re really good at, and be honest with yourself, which is not something this industry is particularly good at.”

Partnerships stall, he adds, in keeping revenue-targeted teams comfortable with overlap rather than at the top, where sign-off is usually easy. Zero overlap is so rare that if it existed, the deal would already have been done.

2,000 advertisers on ChatGPT, with no guarantees

The OpenAI partnership is the test case. Generative and agentic AI have pushed up the value of high-quality data, and when OpenAI decided to move into advertising, there was a live question over whether it would bring the transaction in-house. It stayed open to purchases remaining with the retailer or e-tailer, and that was when the two companies started talking. “They’re not specialists on advertising,” Fischli says. “So we bring that data set, which makes the targeting better, helps with creative, and the measurement will be moving towards much, much better. It helps them grow and kick-start their advertising business. They bring a lot of the tech, and we bring a lot of the data set.”

Criteo is not OpenAI’s only partner, but it was the first advertising technology partner in the ChatGPT ads pilot in March and, according to Fischli, remains comfortably the largest. “There’s a first-mover advantage, and we have by far the biggest. We’ve just announced we have 2,000 advertisers live on OpenAI. This is significant in just a few weeks.”

He describes the relationship as close, particularly with OpenAI’s advertising team, and insists the commercial story has barely started. “Our joint clients haven’t seen anything yet on the true value that this will have. It’s very early days, because the measurement piece, we’re only starting to implement.”

The pace of change makes longer bets pointless. On a travel panel that week, he and fellow panellists compared their predictions from a year earlier. “What was the delta? Oh my god, 75%. Scrap everything.” Criteo’s insurance is integration: an advertiser can compare the value of the OpenAI part of a campaign against the Facebook and retargeting parts in one place. “I won’t quite call it a moat, but it’s definitely still a bit of protection in the short term.”

“One of the things I’m usually talking about is, is the age of the moat over? When digital advertising first started, everyone wanted to create a moat. Some people managed to do that, and it worked well for a number of years. Given the speed of innovation, I don’t know whether the age of the moat is still here. What will win, in my opinion, is not necessarily the people who create the best moat, but the people who remain the most adaptive.”

Six months to democratised MCP

His most testable claim concerns agentic buying. In May, Criteo and Dentsu announced what they call the industry’s first MCP campaign: a Dentsu trading agent orchestrated it end-to-end for a French client through Criteo’s Model Context Protocol, with no one logging into a Criteo interface. “The results are fantastic,” Fischli says. “I think we’re going to see in the second half of this year a positive explosion of the usage of this. Over the next six months, we will see the democratisation of MCP.”

He is dismissive of the wider industry’s claims on this front. “A lot of people are telling everyone they’re investing in this. The amount of people that have done something that truly works is very, very small.” Once democratisation arrives, he expects the industry to catch up within a further six months. “Agent-to-agent will be a true reality when we talk next year.”

Criteo’s protection rests on data exclusivity, and its newest partner sits closer to the consumer conversation every month. On Fischli’s own telling, nobody in this market can answer that question for more than six months at a time.

Andy Burnham became UK Prime Minister without a contest, a challenge or a televised debate, and, as the Spectator noted with some alarm, without taking press questions at either of the set-piece moments of his rise: his by-election victory and his “No 10 North” policy speech. For most incoming prime ministers that would read as ordinary media management. For this one, the national press is entitled to wonder whether it is a preview.

Burnham is the most committed supporter of press regulation ever to enter Downing Street. In 2016, as shadow Home Secretary, he called Leveson Part Two “non-negotiable” and wrote to ministers demanding it proceed, telling Radio 4: “The Government’s only done half the job.” This year, before the leadership crisis broke, he said of the abandoned inquiry: “I haven’t given up on that”, adding that he would revive it if he “had the influence”.

He now has the influence.

The motive is foundational, not tactical

Most politicians’ views on press regulation are positions. Burnham’s are biography. His politics were forged in the Hillsborough campaign, and his view of the national press, above all The Sun, was forged there too. The campaign that made him, the fight to overturn a false narrative that a newspaper printed and defended for decades, is not an episode he can be lobbied out of. When he linked press reform to Hillsborough in 2016, he was describing the foundation of his public life.

That distinguishes him from every predecessor who inherited the Leveson question. Cameron commissioned the inquiry and shelved its second part; the May government formally abandoned it in 2018, over Sir Brian Leveson’s own objection, and saw off a High Court challenge to that decision; Johnson, Sunak and Starmer had no appetite to reopen it. Burnham arrives with the opposite problem: a personal commitment made repeatedly and in public, and a government to run.

The battle lines were drawn before he got the job

The lobbying began weeks before the removal van reached Downing Street. Hacked Off has been open about its hopes; Hugh Grant has pressed the case in person. On the other side, the Telegraph has already run the “war on journalism“ counter-offensive: a panic its critics note is not disinterested, given the paper’s deputy chairman sits on the board of the company that funds IPSO, the industry’s self-regulator. An unnamed industry source has warned that if the government’s media green paper were “hijacked to advance a press regulation agenda”, it would undermine the stated aim of combating misinformation.

That green paper is where intention will first meet timetable. Lisa Nandy’s consultation, launched on 23 June with proposals to make “trustworthy” news prominent in social feeds, closes on 31st August landing directly on the new prime minister’s desk. Prominence for trusted news could be commercially significant for national and local publishers alike. It could equally become the vehicle for a standards agenda, depending on who defines trustworthy and how. The Press Recognition Panel has already welcomed the paper in terms the industry will not enjoy. Whether Nandy keeps DCMS in this week’s reshuffle, and who gets it if she does not, is not a detail this time.

Subscribe now

What he could do, and what it would cost

The instruments vary in price. Section 40 of the Crime and Courts Act, the costs-shifting provision publishers spent a decade fighting, was repealed by the Media Act in 2024; restoring it would require primary legislation and a bruising parliamentary fight. Reconvening Leveson Part Two, by contrast, is a decision, not a bill. An inquiry into historic press conduct could be restarted by a government that wanted one.

But wanting one and affording one are different things. Burnham takes office with Labour only just drawing level with Reform after a year behind, a cost-of-living agenda he has promised to put “front and centre”, and limited capital to spend on a fight with the national press that most voters stopped following in 2012. Publishers betting on his restraint has the political arithmetic on their side. The arithmetic, though, is the only part of this that moves.

The paradox: their examiner may also be their defender

The prime minister most likely to reopen the question of how newspapers behave is also the one most likely to defend what they own. Publishers could find their conduct examined and their copyright protected by the same government. No occupant of No 10 has offered them both at once, and how the industry plays it, which fight it picks, which gift it accepts, will be a story in itself.

Most people entered this industry because of their creativity. The uncomfortable question, ten months into Ndidi Oteh’s tenure as Chief Executive of Accenture Song, is how much room is left for it. Her answer, delivered in a recent joint appearance with Creative Chairperson Nick Law, was unequivocal: plenty, but only for organisations that treat creativity as a discipline rather than a department. Disciplined creativity, applied across every part of a business, is, in their telling, the difference between being valuable and being replaced.

That framing is worth taking seriously, not because it is novel (consultancies have been promising to operationalise creativity for a decade) but because of who is making it and at what scale. Accenture Song is, by revenue, the largest creative business in the world: $20bn in its 2025 fiscal year, up from $12.5bn when David Droga took charge four years earlier. When its two most senior figures set out a doctrine, they are describing how a meaningful share of global marketing, commerce and experience spend will be organised.

The company, not the CEO

Oteh took over from Droga on 1 September 2025, after 14 years at Accenture and, from late 2023, a stint running Song’s Americas business. She has been careful not to position herself against her predecessor, describing him as a mentor and team builder with a habit of taking on what everyone else called impossible. Droga has moved upstairs to become Accenture’s Vice Chair; the succession was orderly, and the strategy is continuous.

Around a decade and a half ago, she recounted, Accenture concluded that technology was reshaping how consumers live, learn and socialise, and that it wanted to own the consequences. It began buying and building creative capability under the Accenture Interactive banner, established in 2009: Fjord, Karmarama, The Monkeys and, in 2019, Droga5, the deal that signalled the ambition was serious. In 2022, with Droga newly installed as Chief Executive, Interactive and its more than 40 acquired agencies were stitched into a single brand and renamed Accenture Song; only Droga5 kept its name. The point, Oteh stressed, was never to assemble a holding company in miniature. Accenture set out to be neither a consultancy with an agency bolted on nor a technology firm that dabbles in advertising, but one integrated business spanning marketing, sales, commerce and service.

That integration is the product. Song’s pitch to clients is what Oteh calls harmonised consumer experiences: the same brand behaving coherently across every touchpoint, all the time, rather than in campaign-shaped bursts. “Technology has always needed creativity, and creativity always needs technology,” she argued. The companies that fuse the two with intention, in her view, are pulling away from those that do not, and the gap is compounding. It is a tidy inversion of the usual anxiety: rather than technology eroding creative value, Song’s case is that technology has made creative judgment scarcer and therefore more valuable, provided it is wired into the whole enterprise.

Read more

Paul McCarthy-Brain hears the same objection so often he has a name for it: nobody has heard of his company. “We’re kind of neverheardofyou.com unless you’ve heard of us,” quips the Founder and Chief Executive of Flip-Pay, the Dublin content monetisation company he launched in 2015. “We operate predominantly off referral. Our sales team is small, and we generate more work than we can handle.”

The anonymity sits oddly against the client list. Flip-Pay’s platform took Mediahuis Ireland from zero digital subscribers in February 2020 to 100,000 across the Irish Independent and the Belfast Telegraph by June 2025. It runs HuffPost’s membership programme, Euractiv’s access systems and Press Gazette’s metered paywall.

In May it powered the Salt Lake Tribune’s removal of its paywall, the most closely watched reader revenue experiment in American local news. This year the company has engaged PR for the first time. “It’s about time we actually told people what it is that we do.”

Micropayments died in the finance department

Flip-Pay’s legal name, International Micro Payment Ventures Limited, records the founding idea. “We actually started off as a micropayments play,” says McCarthy-Brain. He had been investigating whether cryptocurrency wallets could finally make small payments work for publishers. Three years of R&D followed. The first phase dug into why every previous attempt had failed; the second, run with News Corp in the UK and US, explored what a working version might look like.

His conclusion was that the industry had misdiagnosed the problem. The real barrier was moving a reader’s identity and entitlements between publishers without friction, and no vendor had cleared it.

The technologists liked it. Its accountants did not. “The CFOs always found it too disruptive. These are businesses predominantly based on recurring monthly and annual subscriptions, growing 2 or 3%. Something which would cannibalise that business would present too much of a risk.”

The BBC came closer, exploring a journalism wallet for small, regular payments. “They got it straight away. But the BBC is slow. The heads of tech and the director generals come and go. We were one of those cycles, and it ended up going nowhere.”

Mediahuis made the pivot pay

The turn came from Independent News & Media, preparing the Irish Independent’s move to paid access under its new owner, Mediahuis. Flip-Pay won the business against Piano, with a brief that reset its direction.

“They said, we love what you’ve done with the tech you’ve built, but would you go and build the gated community, for the monthly and annual subscriptions? Now we can do everything at the micro level. A day pass, an individual article, an e-paper, all the way through to monthly and annual subscriptions.”

That relationship still defines the pitch: “Really customised boutique implementations for very innovative publishers. That’s where we’re making our bread and butter.” When a client sets what McCarthy-Brain calls a North Star objective, Flip-Pay builds the feature, then folds it back into the platform for every other publisher.

For Euractiv, the Brussels title he describes as “basically the European equivalent of Politico”, that meant a European Parliament integration. Officials log in with EU credentials and get automatic access to pro-level content.

Flip-Pay would rather acquire than be acquired

The sector around Flip-Pay is consolidating fast. Sophi went to Mather Economics in 2023 and Poool to Darwin CX in March. “We’re privately funded. We have no interest in selling. We get loads of venture capital and private equity queries monthly.”

The company went through what he calls a growing-up phase around 2022, taking on ISO, SOC, and NIST certifications. “We stopped being a startup and started being a proper company. We may acquire something, but we’ve no current intention of selling.”

The unglamorous end of the business is growing fastest. In 2024 an introduction from California’s Press Democrat led to an alliance with Teak Systems, pairing Flip-Pay’s digital stack with Teak’s print circulation tools. The target is American publishers on legacy systems such as Naviga, in his words “tech which was developed before the dinosaurs roamed the earth”. That combination, he says, is “growing like crazy” in the US, with Latin America following.

A recurring donation is a subscription by another name

Flip-Pay’s move into donations began with a customer request. “We stepped back and looked at our platform. There’s not an awful lot that’s different between donations and subscriptions. If it’s recurring, basically it’s a subscription. The next question was, how could we deploy paywall tech into the donations flow?”

That means pointing the machinery of dynamic paywalls at the appeal itself: cohort analysis, metering and content-level context. A story about the US government and women’s health services, read by a largely female audience, can carry an ask tied to that coverage.

McCarthy-Brain gives the category’s pioneer qualified credit: “The Guardian is the daddy of this. I think they built it themselves, and it’s quite a blunt instrument.”

His other rule is that the transaction must never leave the publisher’s brand. “The worst thing a publisher can do is link out to a third party. You’re taking your money and giving it to somebody else. It should be invisible to the user.”

“You can’t just stick up a thing saying give us money, donate now. It just doesn’t work. Nobody will. The tech is fine. We’ve got the tech; it works. The hardest period is coming up with the narrative. Why should you support this publisher?”

Converts, he argues, are worth more than the payers they replace. “Subscribers are transient for the most part. If you can sell them on why what they’re doing matters, you’ve got a much stickier subscriber. It’s a different model, but the mechanics are all the same.”

HuffPost tested the thesis at scale, launching voluntary contributions with Flip-Pay in 2023 and evolving them into a tiered membership programme in January 2025.

Salt Lake did its homework before the gamble

The Salt Lake Tribune is the harder case. The Utah Daily has been under nonprofit ownership since 2019. In mid-May it took down its paywall and asked its 32,000 digital subscribers, roughly a third of its revenue, to become voluntary members instead.

Press Gazette framed the move as a gamble with that third. McCarthy-Brain rejects the framing. “They’d done their homework. They tested the waters. They didn’t just jump into this. They engaged with customers. A high percentage said they would convert.”

The conversion was heavy engineering: migrating subscriber data off a legacy platform, rebuilding onboarding journeys and satisfying “a lot of really specific non-profit accounting rules and regulations that you absolutely must follow.”

Early behaviour supports his stickiness argument. “People who may never subscribe are now deciding to contribute.” Lapsed payers are not written off either; they “go into your standard dunning flow, like you would a subscriber, and they can be targeted later.”

Flip-Pay now hopes to work with the American Journalism Project to replicate the model across the US local news landscape. On the Tribune itself, he hedges: “Time will tell. Once we’ve done a couple of years with them, we’ll get some really good solid data.”

The next subscriber may be a software agent

Collapsing Google referral traffic is doing Flip-Pay’s selling for it. His shorthand for what publishers on legacy paywall tech is telling him: “This was fine six months ago, but now what we need is every single bell and whistle available to us at a better price.” “Just having a paywall up is not good enough these days. You need dynamic elements. You need intelligence.”

On AI answer engines, his assessment is blunt. “At the moment, the only tools you’ve got available to you are to block them or to license, and that is pretty much it.”

The bigger shift, he believes, is agentic purchasing: an AI assistant instructed to find the cheapest route to a title’s journalism and transacting through a programmatic interface. That would bypass the A/B-tested checkout journeys the industry has spent a decade refining. “The content will be consumed in a different way, but there will still be the need for news organisations to tell us what’s going on. If that stops, the AI agents have nothing to read.”

AI has also compressed Flip-Pay’s own build cycles (“When you have an idea, we can prototype and iterate in days”). It is feeding retention too: a cancellation call can be scored in real time against lifetime value and engagement, with a custom offer generated on the spot.

In September, Flip-Pay will launch an unnamed AI product with Press Gazette at the Future of Media Technology Conference, born from a call with Editor-in-Chief Dominic Ponsford. “It started with a telephone call, and it spiralled into an enterprise thing.”

The growth has a lopsided shape, and McCarthy-Brain knows it. When his principal investor asked before the summer what could hurt the business, his answer was concentration. The Americas are expanding fast, and Europe is not keeping pace; a move by President Trump against the dollar or the wider US economy could turn that strength into an exposure. So, after a decade of being recommended in private, Flip-Pay is setting out to make itself known in public, in a market where the vendors it wants to displace have been advertising for years. Expect to be hearing more.

Subscribe now

On Wednesday afternoon, while much of the country was preoccupied with England’s World Cup semi-final against Argentina, the Advertising Association gathered MPs, ministers and industry leaders at the House of Commons to mark its 100th year. Damian Hinds MP hosted, opening with a tour of advertising’s place in British cultural life, from the Gold Blend couple (Nescafé’s long-running coffee romance) to Nick Kamen’s 1985 Levi’s 501 launderette and what he called those awful meerkats (Compare the Market), which drew gasps from the audience, before handing over to AA President Andria Vidler.

Culture Secretary Lisa Nandy sent her apologies; she was in Atlanta for the match. And as Times Radio’s Kate McCann reminded the room, Andy Burnham is expected to become prime minister by the end of the week. A trade body celebrating a century of continuity was addressing an industry, and a political class, in the middle of a rupture.

Trust gets promoted from reputation defence to growth strategy

Vidler, Chief Executive of Allwyn UK and Advertising Association’s President since March 2025, used her speech to reframe the association’s founding principle. Trust in advertising sits at a five-year high, she said, crediting the Advertising Standards Authority (ASA) awareness campaign and the media owners who fund it. The industry the AA serves has also transformed in scale, from 300,000 businesses using advertising 25 years ago to 3.6 million today.

Her argument was that the industry has spent years treating trust as a soft measure for defending brand reputations and should now treat it as a driver of effectiveness. Trusted brands, she argued, grow faster, command price premiums and keep customers loyal.

“We believe that we’re seeing a real shift from the attention economy to the trust economy, where the advertiser that wins is the one that really builds and reinforces trust in every aspect of what it delivers,” This shift to the trust economy means treating trust no longer as a soft measure, but as a hard commercial metric. The king of KPI’s”

Andria Vidler, President of the Advertising Association and CEO of Allwyn

AI adoption is the accelerant. Mobile phones took 16 years to reach 100 million users, she noted; the Internet seven, Facebook four and a half, TikTok nine months and ChatGPT two. With the public using AI to make choices at that pace, trust becomes the basis on which brands are chosen at all. Her prescription for the centenary year: continued backing for the ASA, a business case for trust built on evidence such as the Institute of Practitioners in Advertising (IPA) effectiveness databank, and a higher profile for the Department for Culture, Media and Sport (DCMS) Online Advertising Taskforce and its work on transparency and protecting children. “Advertising that builds trust must be central to every brief,” she said.

Subscribe now

Government backing arrives with a fraud problem attached

Ian Murray, Minister for Creative Industries, Media and Arts, followed with the government’s side of the ledger. Advertising was named a frontier subsector in the creative industries sector plan. The government supported a trade mission to Cannes Lions, where he said UK work took more than 80 awards. Murray jointly chairs the Online Advertising Taskforce, paying tribute to industry Co-Chair Mark Lund.

“There is no place for fraudulent advertising,” Murray said, pointing to the consultation Ofcom published last week on the ad fraud duty under the Online Safety Act and urging the room to engage with it before it closes. The taskforce’s Ad Fraud Working Group, involving the IAB (Interactive Advertising Bureau), DCMS and the Home Office, has reported back on phase one of its work; phase two, developing solutions the industry can apply to the programmatic supply chain, starts now.

Murray also placed the TV green paper published in June, the local media strategy and the ongoing radio review within the same argument: sustainable, trusted media underpins the platforms advertisers depend on. “It’s in all of our interests to put trust at the heart of everything we do,” he said.

The audience, as Times Radio describes it, is in no mood to be reassured

Stig Abell and Kate McCann, Times Radio’s breakfast presenters, read out the mood of their listeners, a self-selecting national focus group that messages the programme in real time. Abell made the commercial stake : “If we don’t have listeners, and if those listeners don’t listen to ads and don’t pay for us, we don’t exist.”

Three themes dominate their inbox. The first is accountability. McCann described listeners spread across ages, regions and incomes, united by unease and by the conviction that politicians answer a different question to the one they were asked. The second, in Abell’s telling, is the establishment: who runs Britain, who is responsible for years of stalled growth, and why anyone claiming authority, whether MP, journalist or brand, should be believed. The third is the economy, sharpened by online calculators speculating about household bills under the incoming government. “The one thing that they hate more than anything else is the feeling that politicians are still lying to them,” McCann said. Her advice, echoed by Abell, was humility: show your work and admit what you do not know.

Stephen Woodford, the AA’s Chief Executive, closed the reception with a toast “to trust in advertising over the next 100 years”, describing trust as close to a universal panacea for politicians, business and the industry alike. That is the tension the centenary leaves open. The AA has data showing trust in advertising rising. The audience described from the same stage has rarely been more sceptical, and by the weekend it will have a new prime minister to test. Its second century opens on that gap: an industry betting on trust just as the public runs short of it.

Related Articles

Whatever else Cannes Lions 2026 was about, it was about AI. “I asked someone what they thought was the main theme. Obviously, this year you hear AI everywhere. I feel like it’s much more prevalent than last year,” Sascha Eder told me when we sat down at the Palais. It was NewtonX’s second time attending the Cannes Lions Festival and a more organised one, he said: a team of four, a diary of client meetings, and a session on synthetic B2B buyers that put him on stage alongside Jamie Cleghorn, Global Head of Customer Practice at Bain & Company and Zoom’s CMO, Kimberly Storin.

Eder is the CEO and co-founder of NewtonX, the New York-based B2B research intelligence company he started after a spell at McKinsey, where the frustration of advising clients without access to decent data became the founding thesis. The company now works with more than 600 of the world’s largest organisations, including Google, TikTok, Salesforce, Stripe, Microsoft, and Coinbase, supplying what he calls hard-to-get information: the verified professional insight that lets an enterprise, a consultancy or a hedge fund make decisions it would otherwise be guessing at.

The bots reached the research panels before the buyers did

The reason NewtonX exists, Eder argues, is that the market research industry has never fixed its supply chain. “For the last 25 years the industry has been rampant with fraud, and that’s being exacerbated by AI. More bots, even more sophisticated fraud.”

The mechanics of that fraud have evolved fast. “These days you have to deal with sophisticated people impersonating on LinkedIn, you have to deal with bots, you have to deal with AI agents that can simulate a real response in a survey that you almost can’t distinguish from a human.”

His answer is a verification stack rather than a single check. The NewtonX Graph indexes over 1.1 billion professionals, enriched over the years with signals such as IT and advertising budgets by industry and company size. When a client wants advertisers with retail budgets above $20 million, the company pinpoints them, recruits them across multiple channels, then verifies them individually: corporate email, phone, social profile age and authenticity, and topic-specific testing on every project. “There’s not one piece of verification that is bulletproof. It’s really the combination of multiple layers that makes it safe.”

The panel model, in which respondents self-report, is where the rot concentrates. On a panel, he notes, people designate themselves as a CFO one day and a CMO the next to qualify for more surveys. Pre-targeting from verified data closes that door.

Cheaper and faster are conditional claims

Verification is the plumbing. Synthetic personas, the subject of Eder’s Cannes session, are what the industry queued up to hear about, and the appetite is easy to understand. They promise research without recruitment: an always-available simulated buyer that answers immediately and at scale.

Eder’s definition carries a condition that most of the market skips. “A synthetic persona is an AI model that is built, especially in the B2B space, on real verified buyers. If you try to sell to an IT buyer that sits in mid-market finance companies, the model learns from those. The premise of synthetic is it’s cheaper, it’s faster, it’s immediate, and you can do much more at scale, but people tend to forget these things only hold true if the model actually works.”

​He is equally deliberate about what the technology is for. “It’s not about whether synthetic research is better than real research. It’s about the use cases where teams don’t use any research at all, where they guess the answer, or wait weeks or months to run a real study.”

​On familiar territory, the pitch is arithmetic. “Instead of launching 100 ads, assessing them and paying $50,000, you can launch 1,000 ads, have the synthetic personas pre-filter them down to 20, launch those, spend $10,000 and have as much or more conviction than the old method.”

​He tells clients to walk away when the concept is genuinely new. “If you say, here’s the new hardware device OpenAI is launching that’s supposed to replace the iPhone that no one has ever seen, and you ask a synthetic audience, that’s not going to work, because it’s a brand-new concept.” The same logic applies to money. A new product line with $100 million to $500 million behind it gets primary research, no argument.

​Agencies may be the buyers most exposed to that trade-off. Demand from them has exploded in two years from a standing start, and for independents priced out of enterprise-grade B2B data, he is blunt about what synthetic offers. “B2B data is expensive, high quality is expensive, but this is one of the areas where synthetic can provide a lifeline. Is it the same as primary research? No. But is it better than these agencies not having access to any data? Yes.”

Subscribe now

Nobody wants to talk about stale personas

Verified data ages, and a persona trained on last year’s market is quietly answering last year’s questions. Eder concedes this is the part of the pitch the industry avoids. “Nobody talks about it, either because they don’t want to do it or because they’d rather not talk about it. The reality is you’ve got to update the personas every few months.”

NewtonX builds proprietary personas for clients and refreshes them roughly quarterly, using what it calls persona shells: behavioural and attitudinal segmentations for specific buyer types, enriched with the client’s own data. Refreshing every three months, he argues, still costs less than running a heavyweight brand tracker.

Validation is done the hard way: the company splits a real audience in two, builds the model on one half, runs live research on the other, and compares the results. “Confidence, depending on the topic and the audience, can vary anywhere from 85% to above 95%, which is extremely high in some cases. I would argue even 55% or 60% would be more than sufficient, because the trade-off is not making a decision based on any data at all, and there your confidence gets close to zero.” The effort is led by a VP of Research and Innovation who previously ran research at Salesforce-owned Tableau.

Half the buying committee never shows up in the CRM

Push on the hardest objection, that B2B purchases are shaped by politics, budget cycles and relationships that never appear in survey data, and Eder does not retreat into the model. “People are irrational. People say one thing and do another.”

His response draws on the multi-year buy-ability research NewtonX has fielded for LinkedIn and Bain, which mapped how enterprise buying committees behave. “You have the buyers everyone thinks of, and then you have hidden buyers, people in the background nobody is aware of, but they influence the decisions. If you’re not aware of it, you’re at risk of losing out.” The published findings put roughly half of decision-making influence with those hidden buyers in procurement, finance and legal. Personas built on that foundation can model committee dynamics to a degree, he says, but only because the primary research identified the dynamics first.

​That insistence on primary data is also what separates B2B from consumer work. Consumer synthetic tools benefit from an ocean of training data; B2B does not. “People say, well, synthetic worked for B2C, so if I train my model on the web, I’m sure I can ask it to be a procurement specialist. That’s where the model doesn’t have the data, but makes up the audience nonetheless, and gives you a confident-sounding answer.”

The replacement narrative did not survive contact with the data

NewtonX’s study with The Wall Street Journal, which surveyed 82 brand marketing VPs on generative AI, went in carrying three hypotheses: that marketers expected AI to replace people, that it was already generating strong returns, and that cost-cutting was the point. “All three were debunked.” Marketers wanted efficiency and better campaigns; the returns were prospective; and the head-count story kept collapsing on inspection. “In many cases the hiring actually goes up when you deploy AI, because you now need new roles. You go from individual contributors to agent orchestrators. The skill sets change, which is also an argument for people not being replaced but needing to be reskilled.”

The business behind the argument is in credit. NewtonX raised a $32 million Series B at the end of 2021, taking total funding to $47 million, reached profitability within two years and is funding growth organically, with around 200 staff, a London office building out its European presence, and a fresh citation from Gartner as a frontrunner among synthetic population and behavioural simulation startups.

Everything Eder sells depends on buyers who ask what the model was trained on, and he is candid that only the sophisticated ones do. The rest of the market is being offered an ever-cheaper way to hear a confident answer from tools trained on the same polluted panel data that created the crisis in the first place. Twenty-five years of fraud sits underneath them, and nothing at this year’s festival suggested the industry is in a mood to slow down and check.

Related Articles