From visibility to recommendations: how earned media is powering AI search

A decade ago, SEO reshaped how brands competed for visibility. Today, we’re entering the generative engine optimisation (GEO) moment where large language models (LLMs) aren’t just retrieving information but evaluating, comparing and recommending brands in real time. And for PR and communications professionals, the implication is stark: this is not an opportunity we can afford to sit out.

Earned media is one of the most influential signals shaping how AI understands and trusts brands. As LLMs evolve from passive tools into active decision-makers, the narratives they absorb from the media ecosystem increasingly determine who gets surfaced – and who doesn’t.

New research from Hard Numbers and Onclusive shows just how real this shift already is. Media performance now has a direct, measurable impact on how brands appear across AI-powered platforms. Brands with a strong earned media footprint are consistently outperforming competitors in AI-mediated search and recommendation environments.

How earned media influences AI search

Our Coverage to Capital: Recommendations in the Age of AI report analysed over 100 global brands across multiple LLMs, including ChatGPT, Perplexity and Gemini. The results were unambiguous.

Using Onclusive’s global media influence score. which combines media volume, reach and sentiment, we found that brands in the top 10% for media influence are recommended 80% of the time. Brands in the bottom decile struggle to reach even 50%.

LLMs are far more likely to recommend brands with strong, visible earned media profiles. Not because the models are biased, but because they are learning from the same ecosystem of coverage, commentary and credibility cues that shape human judgement.

Far from diminishing PR’s value, AI is reinforcing it.

Why sentiment now carries more weight than ever

If media presence gets a brand into the AI’s field of view, sentiment often decides the outcome.

Our research shows that brands in the top 10% for positive media sentiment are nearly twice as likely to receive low concern scores from LLMs – a useful proxy for how “safe” or credible they appear in procurement-style evaluations.

But the most eye-opening finding is this: brands with highly positive sentiment are three times more likely to win head-to-head LLM comparisons than those at the bottom of the scale.

Ask an AI, “Which is better, brand A or brand B?” and the answer increasingly reflects the balance of positive and negative narratives that brand has accumulated over time.

As LLMs move towards autonomous agents – capable of shortlisting suppliers or influencing purchasing decisions – these gaps stop being theoretical. They become commercial.

And unlike traditional news cycles, AI doesn’t forget. Negative stories have a long digital half-life, continuing to influence concern scores long after they’ve disappeared from headlines.

What PR teams should focus on now

If earned media plays a material role in how AI perceives brands, PR teams need to adapt but they don’t need to reinvent the wheel.

The fundamentals still matter. High-quality earned coverage should be the starting point. Volume has its place, but well-circulated, credible, positive coverage matters far more in shaping AI judgement.

Diversity of presence is also critical. LLMs ingest a wide range of sources including business press, trade media, expert commentary and thought leadership. The broader and more authoritative that footprint, the stronger a brand’s overall AI signal becomes.

Reputation management takes on new urgency too. In the GEO era, there’s no chip paper for bad news to disappear into. Addressing issues early is the most effective way to avoid long-term algorithmic drag.

Finally, measurement needs to evolve. What gets measured gets managed – and what gets managed gets budget. While GEO metrics are still emerging, teams can start by tracking AI-referred traffic, monitoring how brands appear in LLM outputs, and analysing how sentiment correlates with recommendations.

AI is already reshaping how decisions are made and earned media is proving to be foundational to this new eco-system. The PR teams who recognise this now, and act with confidence and curiosity, will be the ones whose brands are trusted in the next era of search.

Darryl Sparey is co-founder and managing director of Hard Numbers which is a CIPR corporate affiliate member.

Find out more about a CIPR corporate affiliate membership
Strengthen your reputation and invest in your team’s success. Join as a CIPR Corporate Affiliate member - open to groups of five or more.