To celebrate PRmoment's next PR Masterclass: The Intersection of PR and GEO we've asked a selection of PR innovators to write about their GEO journey and share how their PR workflow has shifted to ensure earned media impacts the LLM results.

For the first time, PR is not just protecting reputation. It is actively shaping how brands are discovered, cited and recommended by large language models.

Over the past year, the conversation has shifted from “what can AI do for PR?” to “what can PR do for AI?”. At the start of 2024, our focus at Third City was on the productivity benefits AI could bring to the agency through our AI taskforce. But it quickly became clear, through conversations with clients, that the most immediate and disruptive impact of AI was on how brands appear and are cited in AI-generated responses.

As tools like ChatGPT entered mass adoption, we saw that large language models are heavily skewed towards trusted editorial coverage. In other words, the core work of PR, earning third-party authority, is now directly fuelling AI-generated answers. In practical terms, this means PR is influencing what LLMs recommend at the moment of consumer decision, a role traditionally associated with search and performance channels.

This shift underpins Generative Engine Optimisation, or GEO, a category we embraced early with the launch of our ‘GEO for Brands’ service in August, amid widespread scepticism. At the time, many dismissed GEO as a category error, assuming it was a rebrand of SEO. But this misunderstands how AI search works.

Unlike SEO, which is largely deterministic, generative engines are probabilistic. The same prompt can generate different answers, tones and sources each time it is asked. A single AI response tells you very little. What matters are the patterns that emerge across many responses. Understanding AI visibility therefore looks less like traditional search analytics and more like quantitative research. GEO, for us, is not about gaming algorithms. It is about ensuring brands are referenced and trusted within AI-generated narratives, just as PR has always done for human audiences.

Recognising early the need for credible probabilistic analysis, Third City made a major R&D investment to develop a proprietary GEO tool, built in close collaboration with our near-shore engineering team. Designed specifically for the probabilistic nature of AI search, the platform runs each user prompt hundreds of times across multiple generative engines. This allows us to surface recurring patterns, quantify the likelihood of brand mentions and identify the sources AI systems rely on most consistently. This sustained investment has underpinned six months of GEO work with global brands including Kia and Dr Martens, and forms the foundation of the insights shared in our interim GEO report.

Our findings show that clear, consistent messaging across multiple reputable titles does more than signal credibility to consumers. It also shapes how AI engines assemble brand narratives. Generative models read broadly and synthesise meaning from a wide range of cues and sources. Controlling that narrative through targeted, joined-up communications is therefore essential. For brands, this consistency translates into greater share of voice in AI responses and increased share of consumer consideration.

Alignment across channels is also becoming critical. The brands that perform best in generative search are those where PR, SEO, social, e-commerce and influencer activity reinforce the same language and proof points. Helping clients achieve this internal consistency will be central to effective GEO strategies.

Optimising for AI search is not about chasing algorithms or producing AI-first content. It is about strengthening the signals generative engines already trust: authoritative editorial coverage, clear and structured owned content, credible reviews and real-world proof points. Brands that perform best are those that answer practical consumer questions clearly and consistently, over time and across channels.

This is PR’s next evolution. Not a departure from the craft, but an expansion of it. When deep media understanding is paired with data literacy and robust auditing tools, PR becomes a measurable driver of AI visibility, not just brand reputation. In an AI-led discovery landscape, GEO makes the value of PR unmistakable.

If this article has inspired you to embrace the earned media opportunities of LLMs, do take a look at the next PRmoment PR Masterclass: The Intersection of PR & GEO.