Discover why LLMs reward credible journalism, expert commentary and structured owned content – and why spray-and-pray digital PR won’t win you AI visibility.
A recurrent belief in SEO circles is that the way to appear more prominently in large language models is simply to “do digital PR”. The logic is that AI systems need references, so if you generate enough coverage, especially with backlinks, you will surface more often.
But that advice collapses on contact with reality. “Digital PR”, as the term has been used over the past decade, does not describe rigorous communications work. It describes a bundle of practices that can range from shallow to outright damaging.
And it is precisely these practices that have weakened media relations, eroded trust with journalists and blighted the reputation of PR as a discipline.
Why digital PR no longer delivers visibility in LLMs
It is important to say this from a position of experience. We are not outsiders throwing stones. We have delivered digital PR ourselves, at a high standard, and have won UK and EU Search Awards for it. We are also working on a number of live GEO briefs.
My agency, PR Agency One, understands what good digital PR looks like and how it supports broader brand building. But the discipline jumped the shark a couple of years ago. The best practitioners have moved into more strategic, integrated roles; the rest have been left with a model that no longer works for brands, journalists or – crucially – the way AI systems evaluate authority.
And for the love of God, please don’t call it “traditional PR”. There is nothing traditional about it, and if you insist on using that phrase I may be forced to stage a minor intervention involving a very pointed look.
Recent examples of shoddy ‘digital PR” across the UK make the point starkly.
How bad digital PR damages media relationships and brand trust
Journalists increasingly complain about spray-and-pray email blasts: irrelevant pitches sent to hundreds of reporters at once, often with the wrong beat, the wrong name or obvious mail-merge errors.
Tech writers receiving press releases about fermented foods; specialists in finance pitched stories about dog costumes; reporters hounded with a dozen or more follow-ups for a single, low-value idea. This is not outreach. It is noise.
In another widely shared episode, a PR agency demanded that The Sun retrospectively add a backlink to coverage the paper had already run – citing, bizarrely, “Creative Commons” as justification.
From backlink obsession to misinformation laundering
Now, we have all asked for a link but this reason was spurious and the reporter removed the agency from her list entirely. A single backlink obsession cost that firm access to a national newsroom. It is the perfect illustration of a worldview shaped by SEO metrics rather than media relationships.
More troubling still was the Press Gazette investigation into agencies using AI-written content and fabricated personas to generate “stories” for link-building. Reporters were pitched pieces about supposed “accident-prone lottery winners”, only to discover that these people did not exist. The agencies involved changed their sender names and domains repeatedly to dodge spam filters. Some newsdesks unwittingly published the material until the ruse was exposed. This is not PR. It is misinformation laundering.
Alongside this sits the conveyor belt of trivial surveys and gimmicky listicles – the “Britain’s most embarrassing habits” genre, pumped out to win quick placements and low-quality links. Much of it is designed for volume rather than narrative.
And at the extreme end, there have been ethical breaches: pitches exploiting tragedies or sensitive news as hooks for irrelevant brands; attempts to pay reporters to run advertorials as if they were editorial; and other tactics that would be unthinkable in any serious communications practice.
These are not fringe cases. They have shaped how “digital PR” is understood by many newsrooms. They are also the reason some practitioners have been made redundant recently, while the better ones have either moved into technical SEO or have begun to learn what proper PR looks like.
What Google and AI research tell us about LLMs and PR
All of this matters because Google has now made the position explicit. A senior executive confirmed that AI systems use all publicly available content in earned and owned media to produce its content. Public articles, expert commentary, lists, explainers, social discussion and well-structured owned content as direct evidence when deciding what and who to recommend.
Our own AI optimisation research reinforces this: more than 75 per cent of the information ChatGPT drew upon came from earned, shared and owned sources shaped by PR professionals, with high-authority journalism by far the dominant input .
Crucially, AI models do not reward gimmickry. They do not treat fabricated personas, trivial surveys or manufactured listicles as meaningful signals. They weight credibility, authority, trustworthiness and public reputation. In effect, they behave like reputation engines.
Which leads to the uncomfortable truth for parts of the digital PR sector: the practices that have defined much of “digital PR” over the past decade are precisely the ones large language models are designed to ignore.
Why proper PR builds the signals LLMs actually value
To be visible in LLMs, brands need exactly what some digital PR teams have never been set up to deliver: real brand reputation. Consistent, high-quality earned coverage. Expert commentary. Narrative development. Structured and transparent owned content. Healthy social discourse. The hallmarks of genuine media relations, not mechanical link-building.
corners of digital PR have treated proper PR as if it were outdated, something slow or old-school, when in reality it is the most modern part of the discipline: integrated, strategic, and reputation-led.
And now Google has made the position unambiguous. The signals that LLMs value most are the ones proper PR creates – credible journalism, expert authority and coherent brand narratives – rather than the low-grade link-chasing that has been allowed to trade under the banner of “digital PR”.
How SEO and reputation-led PR work together for LLM visibility
None of this is an argument against SEO. Far from it. High-quality technical SEO, well-structured on-site content and properly planned information architecture are essential for LLM visibility.
Good SEOs understand how models interpret structure, context and clarity. They are the ones who make websites intelligible to machines. My issue is not with them – it is with the shoddy end of digital PR that confuses link-building with reputation, and volume with value.
If you want LLM visibility, you need proper PR
So the message is clear. If brands want to appear prominently and accurately in LLMs, they do not need “digital PR”. They need proper PR. And if the industry wants to reclaim its reputation, it is time to say so plainly.
