Earned media drives 84% of citations in AI-generated answers, according to Muck Rack's "What Is AI Reading?" report. Paid media, by contrast, accounts for less than 1% of those citations. That gap signals a structural advantage for PR and communications teams - the very people who build relationships with journalists, shape narratives, and earn coverage - as AI platforms become a primary distribution layer for brand information.
The report has already reshaped how the PR industry understands AI's sourcing habits. But the real work is turning that insight into action. Building visibility in AI search requires applying PR instincts to a new set of channels and tracking methods that go well beyond traditional media monitoring.
Where comms teams should focus their AI visibility efforts
The sources AI platforms cite are broader than a standard media list. Comms leaders need to optimize brand mentions and narratives across traditional media, emerging formats like podcasts and newsletters, social platforms including Reddit and YouTube, review sites, and customer forums. That diversity demands a team structure that mirrors the modern media mix.
Allison Worldwide runs its comms services through a single Media + Influence team. The group conducts GEO audits that produce lists of top-cited publishers, social and forum pages, and independent creators - all assessed within a client's specific industry and the prompts they want to be known for. Those lists then feed directly into content, media, partner, and creator strategies.
Why volume tracking falls short in AI search
Counting citations tells you how often a brand appears, not how it appears. AI platforms synthesize information from multiple sources, which means narrative quality, consistency, and sentiment matter more than raw mention counts. Chasing volume across the infinite variations of possible prompts is also a losing game.
Michael King framed the shift bluntly in his 2025 analysis of Google's AI Mode: "The future of search is probabilistic, the past was deterministic." Because large language models have semantic sophistication, brands can focus on broad topical categories instead of individual keyword strings. Strategic prompt tracking then extrapolates how a brand shows up across those categories.
At Allison, teams use curated prompt lists tied to specific category domains. For each benchmark set, they measure inclusion rates, citation frequency, share of answer, average mention ranking and density, and response accuracy. Together, those metrics paint a fuller picture than volume alone. Custom monitoring agents can also surface which prompts customers actually use over time, revealing gaps and opportunities in how a brand appears for the topics that matter.
Narrative density as the real GEO metric
When a search platform returns one aggregate answer instead of a list of links, the goal becomes what Meltwater calls cross-channel "narrative density." A curated prompt set makes it possible to see whether a brand's story is resonating consistently across a topic area - and where it is not.
That analysis drives the roadmap: which priority topics have significant citation gaps, and what do the current AI answers reveal about how to close them. The next step is identifying which publishers or authors carry the most influence in those answers, and what types of content get cited. Those questions form the foundation of a GEO playbook for the modern media mix. For teams building expertise in this area, resources on AI for PR & Communications and structured paths like AI for Media Relations Specialists offer practical starting points.
Why this matters for PR and communications professionals
AI search is not replacing earned media - it is amplifying its importance. The 84% citation figure means that the relationships PR pros build and the stories they place are becoming the raw material for how AI platforms describe brands. But that advantage only holds if comms teams adapt their measurement to track narrative quality, not just mention quantity, and expand their outreach to the full range of sources that AI platforms actually cite. The teams that treat GEO as a narrative discipline, not a volume game, will control how their brands show up when an AI answers.
Your membership also unlocks:
