AI Is Changing Brand Discovery. PR Needs to Change With It

The public relations industry has spent much of the past 20 years determined to prove its commercial value. We have moved from column inches to impressions, engagement, share of voice, and reputation.

The measurement of success has become more sophisticated, although anybody who has worked in PR for long enough knows that a large number on a coverage report does not necessarily mean very much. AI could change this debate in ways the PR industry has barely started to consider.

People are increasingly using ChatGPT, Claude, Gemini, Perplexity, and Copilot to research companies, products, and markets. A potential customer looking for a payments platform, cybersecurity provider, or construction technology company can now ask an AI platform which businesses it should consider and receive an answer within seconds.

For a brand, the question is obvious. Are you in that answer?

This is quite different from the search behaviour marketers have spent the past two decades trying to influence. Google presents a series of links and leaves the user to investigate them, while a Large Language Model (LLM) does much of the investigation first and presents its version of the answer. Where it gets its information from, therefore, matters.

PR now has another audience

Companies are very good at telling the world what they think about themselves. Websites are filled with claims about innovation, leadership, expertise, and market-leading products, most of them carefully written, optimised for search, and approved by several members of the management team.

AI systems have access to all of this, but they can also find what other people have written about the company. That distinction should matter to anyone working in PR.

A piece of coverage in a respected publication has always had a life beyond the readership on the day it appeared. A journalist has reported the story, an editor has published it, and the article becomes part of the public record surrounding the business.

The same applies when a company’s research is reported by the media, an executive is interviewed about an industry issue, or a bylined article appears in a specialist publication. Those articles are now also being read and interpreted by machines.

Research from Muck Rack’s Generative Pulse, which analysed more than one million citations appearing in AI-generated answers, found that 84 per cent came from earned sources. These extend beyond journalism to include academic, government, NGO, and other independent third-party content, but the finding is significant for PR. AI platforms are drawing heavily on information published outside a brand’s owned channels when deciding which sources to use and cite.

For the PR industry, that creates a new reason to care about where coverage appears, what it says, and the subjects around which a client builds genuine authority.

Coverage may have acquired a longer shelf life

There has always been a slightly artificial end point to a PR campaign. The coverage is secured, the report is sent to the client, and everyone turns their attention to the next story. The internet has already made that approach dated.

An article does not disappear when the newspaper goes to recycling or this month’s trade magazine is replaced by the next edition. It remains online, searchable and capable of being found years later, while LLMs give that archive another use.

Imagine a payments technology company spends three years building a credible media profile around data security. Its executives are regularly quoted by specialist journalists, the company publishes original research which receives media coverage, and its views appear in interviews, news stories, and opinion articles across publications respected within the industry.

Compare that with a competitor which has invested heavily in its website but has generated little independent coverage.

Both businesses may have perfectly good products, but the information available about them across the wider web is very different.

This does not mean the first company will automatically appear above the second in ChatGPT, Claude, or another LLM. Anyone promising that level of certainty does not understand how unstable and opaque these platforms remain.

What can be said is that the first company has created a much broader collection of independent information around its brand and expertise. That has always been good PR. It may now be good AI Visibility as well.

We are seeing it ourselves

Mixology has been looking at this through the work we have been doing with our partners Blue Train Marketing, a digital marketing agency specialising in the fintech and payments. Blue Train works across search, website performance, content and optimisation, helping companies become discoverable and deeply understood online. Our role is focused on PR and earned media, building the independent third-party profile around the brand.

Analysis from one client programme is already showing why this matters.

During a period in which the client’s AI visibility improved significantly across four international markets, its research and editorial content appeared among the sources supporting AI answers, alongside independent articles and industry coverage. PR alone did not create the improvement, but the presence of earned media among the sources being accessed by AI deserves attention.

A different way to think about PR value

AI visibility will not replace conventional PR measurements. Nor should it. Coverage which reaches a prospective customer, changes an investor’s perception, or prompts a journalist to approach a company has obvious value whether ChatGPT sees it or not; but AI visibility does offer another measure of influence.

If a client wants to become recognised as an authority in a particular market, agencies can now look at how that company appears when relevant questions are asked across LLMs. They can examine which competitors appear, which sources are being cited, and where the client is absent, then repeat the same exercise over time.

This has large implications for PR strategies. A campaign which produces dozens of disconnected mentions may generate impressive coverage numbers but leave little coherent evidence of what the company knows. A sustained programme built around a handful of commercially important subjects can create a much clearer body of expertise. Good PR practitioners have understood this for years. AI provides another way of seeing the result.

An opportunity for the PR industry

Much of the discussion about AI and public relations has centred on what the technology will automate. Drafting, research, transcription, monitoring, and basic content production are obvious examples, and agencies will use AI to perform all of them faster.

But we think there is another side to the story which is far more interesting.

AI systems need credible information, and businesses want to be visible when those systems answer questions about their markets. Independent editorial coverage is a potential source for such information, but generating that coverage remains difficult.

You still need an idea worth writing about.

Research must reveal something interesting, and a company needs an opinion which isn’t already shared by everybody else in its industry. Somebody has to understand why a journalist would care and persuade them to cover it.

AI can help with the process, but it cannot manufacture independent editorial endorsement.

That allows PR agencies to broaden the commercial value of what they already do. It will require a better understanding of search, LLMs, and the sources influencing AI responses. Agencies will need to measure more than press coverage, and clients will expect them to understand how their reputation exists across an information environment which is increasingly being read by machines.

PR has always been about influencing what the world knows about a company. The definition of “the world” is changing.