The Daily Mail is testing an advertising proposition built around a new marketing question: how often does a brand appear when an AI assistant answers a customer’s query? Rather than selling exposure only on its own pages, the publisher is exploring whether its branded content can influence which companies get cited in AI-generated answers.
Carly Steven, the publisher’s director of SEO and editorial ecommerce, described the work as “a new potential revenue opportunity” during an industry event in London. A formal commercial offering is planned for early 2027, but the current program remains experimental. That distinction matters: improved visibility in an AI answer is not the same as demonstrated sales impact.
Table of contents
Jump to each section:
- How the Daily Mail tests AI-answer visibility
- Why publishers see a commercial opening
- The measurement problem behind citation gains
- What marketers should know about publisher-led GEO
How the Daily Mail tests AI-answer visibility
The workflow begins with consumer questions. A team identifies prompts someone might use to research an advertiser’s product or service, repeats those queries across AI platforms, and records both the answers and the sites cited. The publisher then assesses whether its own content is likely to be useful in that topic. Where the conditions look promising, it develops tailored branded content and continues monitoring the same prompts.
12 to 15 GEO tests are underway across the United States, United Kingdom and Australia. The tests are intended to inform a commercial offering planned for early 2027, according to Carly Steven’s industry-event comments.
A crucial part of the method is repetition. Steven noted that AI systems can return different sources in response to the same question at different times. The team therefore looks for a stable baseline rather than treating a single favorable answer as a campaign result. It also screens for topics where government or medical organizations dominate citations and a publisher may have little realistic influence.
The mechanism is different from buying a sponsored placement in a search-results page. The Daily Mail is not claiming control over an AI model or guaranteeing inclusion in its responses. It is testing whether relevant material published on an established media domain can become a source those systems choose to reference.
A citation opportunity is not a guaranteed placement. That is the essential distinction for a marketer considering GEO services built around third-party publishers.
Why publishers see a commercial opening
Publishers already have commercial relationships with advertisers, branded-content production teams and web properties that AI systems may cite. The Daily Mail’s experiment asks whether those assets can support a paid service focused on off-site brand discovery, not merely traffic back to the publisher.
It is not an isolated experiment. Future began offering GEO services earlier in 2026 using tailored commercial content across its specialist media properties. Ziff Davis has also reported GEO clients, while Forbes is exploring consultative services that use citation analysis around its Vetted product-review content. The maturity of these offers differs, and public evidence does not establish that they produce comparable business outcomes.
| Publisher | Reported approach | Commercial stage |
|---|---|---|
| Daily Mail | Prompt benchmarking and tailored branded content | Testing; formal product planned |
| Future | Publisher-led GEO content campaigns | Commercial service already offered |
| Forbes | Consultative citation analysis and partner content | Offer under development |
The strategic opening for a publisher is not simply that AI search exists. It is that advertisers may need help understanding which external sources already shape the answers their customers see. A media company can offer content production alongside that diagnosis.
But it also raises an important editorial and commercial boundary. Publishers must distinguish branded material from independent journalism, and marketers should not assume that a recognizable media name will be cited by every AI system or for every product category.
Publisher authority may create an opening, but relevance still determines whether that opening is usable.
The measurement problem behind citation gains
The Daily Mail has shared an early example in which an unnamed personal-finance advertiser became visible in more of the AI prompts under examination after tailored content was produced. The result is directionally encouraging, but the brand, sample design, monitoring period and control conditions were not disclosed. It should be treated as a reported pilot outcome, not proof of a repeatable performance rate.
The more difficult question starts after the mention. A potential customer might read an AI answer, remember a recommendation and later search directly for a company. That path may never appear as an AI referral in standard analytics. Conversely, a citation that produces no meaningful action can look successful if the campaign is judged only on answer appearances.
Steven described an example of a brand using a customer question about discovery sources to supplement referral analytics. That illustrates the attribution gap, not a universally applicable conversion formula. Customer recall can help reveal otherwise invisible journeys, but self-reported attribution carries its own uncertainty.
AI visibility can be measured before its commercial value can be reliably attributed. That makes campaign evaluation a matter of matching evidence to objectives: answer presence, accuracy, qualified traffic, brand recall or eventual customer action are different outcomes.
For marketing teams, this is a reason to be precise about claims. A test that improves source citations demonstrates a change in what was observed on monitored prompts. It does not automatically demonstrate incremental revenue, persuasion or a lasting improvement across all AI platforms.
What marketers should know about publisher-led GEO
The Daily Mail’s work puts a familiar marketing relationship into a less familiar environment. Instead of beginning with a publisher’s audience or a placement on its site, a marketer may begin with the questions a buyer asks an AI assistant and work backward to the sources that shape the answer.
Define the discovery problem. A useful brief starts with relevant buyer questions and the answers currently shown, not a broad promise to increase AI visibility. The publisher’s tests are organized around precisely those prompt sets.
Separate citation share from sales impact. More frequent mentions may be valuable for awareness, but the finance example does not establish a general conversion lift. Decide which outcome the campaign is intended to influence before treating visibility as a result.
Account for the category. Topics dominated by official or specialist sources may offer less opportunity for publisher content. A partner’s domain strength alone is not evidence that it can change results for a specific query.
Ask how observations are repeated. AI-generated answers are variable. A credible test needs comparable prompts, more than one observation and a transparent account of what changed.
The larger shift is that a publisher’s value to an advertiser may extend beyond the people who visit its website. Media content can also become an input into answers assembled elsewhere, although neither citation frequency nor commercial returns are guaranteed.
That possibility creates a new form of collaboration among publisher sales teams, content specialists, search practitioners and brand marketers. Whether it becomes a durable revenue category will depend less on the novelty of GEO than on whether participants can explain, verify and price the outcomes they deliver.
