Daily Mail Runs 12-15 GEO Tests Ahead of AI Visibility Product Launch

By James Calloway

The Daily Mail has between 12 and 15 generative engine optimization (GEO) tests running across the U.S., U.K. and Australia, and plans to turn its AI visibility offering into an official product early next year.

Carly Steven, the publisher's director of SEO and editorial ecommerce, disclosed the figures at Digiday's Media Strategies London event last week. She said the tests aim to work out what actually moves the needle before commercialization.

"We look at this as actually quite an exciting new potential revenue opportunity – not just a defensive play, but [we're] actually starting to think there is a new area for the business," said Steven. "Brands are coming to us already and saying, 'we want to be in these spaces [AI surfaces]. We know that our customers are here. How can you help us?'" she added.

How do the tests work?

The process starts with identifying the prompts consumers might use when researching a brand's product or service. The Daily Mail team tracks those prompts across AI platforms over time, identifies which sources get cited, and, where there's room, creates bespoke branded content to lift the brand's visibility.

Steven stressed the need for long measurement windows. "We know LLMs are probabilistic," she said. "So if I do the same search today and I do it a few hours later, maybe I'll get different results. So we tracked that over a very long period of time to try to get a really solid benchmark to understand where that brand sits within this topic, and crucially, what are the types of sources that are being cited."

The numbers back the approach. In one early test for an unnamed personal finance brand, visibility across target prompts rose from about 50% to nearly 75%, and the content was cited for most tracked prompts.

The opportunity isn't universal, though. Where government or medical sites dominate citations, there's less room for publisher-led branded content. The opening appears when publishers make up a significant share of citations and a brand has little visibility — the team then builds tailored content and tracks whether it improves the brand's presence in AI answers.

How do you prove AI visibility pays?

Attribution is the hard part. AI recommendations don't always produce direct clicks — someone may see a brand mentioned and search for it later. "There's a lot of awareness that there is a gap in the attribution model. It's a bit like putting an advert on a billboard or in a print newspaper. It's how do we measure those things?" said Steven.

Some brands are going back to basics, adding site forms asking customers how they heard about them. One brand Steven cited found Claude accounted for just 0.1% of referrals in its analytics — but roughly 11% of conversions after reviewing customer responses. Referral data alone, in other words, can badly undercount AI's impact.

Internally, the GEO pitch took time to gain traction, partly because the team was still working out how to explain it. Interest grew once brands started approaching the commercial team. Building the offer required closer collaboration across commercial, commerce, native advertising, product and editorial teams — and any new work must protect the publisher's IP and journalism value, including deciding when to unblock access to specific campaign content.

Who else is selling AI visibility?

The Daily Mail isn't alone. Over the last six months, productizing AI visibility has become one of the most talked-about topics inside publishers. Future's GEO product has more than 30 clients and secured renewals. Ziff Davis says it has multiple GEO clients.

Rob Bradley, group svp of digital advertising and operations at CNNIC and Warner Brothers Discovery, told Digiday the topic now comes up in most commercial meetings. "All brands and CMOs, in particular client directors, are talking to us about how they show up and how we show up and how we can work together," he said.

Forbes is taking a consultative route. Chief innovation officer Nina Gould, speaking at the same London event, said brands had approached Forbes for help appearing in AI search, and the publisher started with its Vetted product reviews: tracking citations, optimizing content and sharing results with partners. That work helped Forbes secure better affiliate commissions, sponsorships and native-ad deals.

Forbes is also exploring standalone or bundled products, including ways to assess a brand's visibility before and after a campaign. Gould called it work in development rather than a finished product — but one that has already changed the dynamic with advertisers. "It's deepened the brand partnerships in a lot of ways because, again, it's consultative, so it feels like a partnership between the two brands, which we really like," she said.

With the Daily Mail's product due early next year and rivals already signing clients, AI visibility is moving from experiment to revenue line — and attribution, not demand, is now the bottleneck.