GEO: PR, a Central Lever for AI Citations

The responses generated by ChatGPT, Gemini, or Perplexity largely rely on external editorial content. Studies attribute 80 to 90% of citations to earned media, while AI monitoring solutions are emerging to identify the sources that shape these responses. Communication departments are reactivating their PR instincts to maintain a presence in these generative engines.

AI Monitoring Tools Isolate Media Influencing Responses

Measuring outcomes in AI interfaces remains limited: these systems track traffic, SEO, or sometimes GEO indicators, but rarely identify the specific sources that shape the responses of ChatGPT or Perplexity. To fill this gap, dedicated solutions are emerging. Cision AI Monitor, for example, offers visibility tracking functions in AI engines, model-by-model competitive comparisons, and identification of the most influential media and journalists on generated responses. Monitoring, a historical skill of PR teams, is thus extending to AI to detect problematic narratives, spot outdated information, and measure the real impact of each action. An integrated PR platform that includes AI visibility helps target the right media and journalists to guide a brand's presence in these responses.

Usage: 58% Turn to AI and 90-95% of Citations Come from Media

According to the Capgemini Research Institute, 58% of consumers now use generative AI tools instead of traditional search engines to obtain information about products or services. McKinsey reports that between 90 and 95% of the citations present in responses primarily come from media and review sites, rather than official brand platforms. The user journey is evolving: the question posed to AIs replaces the typed query, and the response takes the form of a synthesized text citing few sources. ChatGPT, Gemini, Perplexity, and Google’s AI Overviews have now become frequent access points for information and brands.

GEO Complements SEO and Strengthens Media Relations Role

Generative Engine Optimization (GEO) complements SEO: instead of aiming for a ranking on a query, it focuses on ensuring that a brand is accurately cited and described within a response. Language models heavily rely on editorial content, and studies conducted by McKinsey, Cision, and Capgemini attribute 80 to 90% of citations to earned media, while purely advertising content is given little consideration. Influencing AIs thus means influencing the media that feed them. In this context, models easily understand press outcomes that are precise, dated, and accompanied by sources. Regular presence in identified media and digital engagement around these publications are perceived as reliable signals by AIs, with visibility in reputable titles carrying as much weight as the quantity of mentions. Preferred methods remain close to those of communication professionals, with a dual advantage: achieving immediate effects and leaving a lasting trace in the models' memory.

Five Operational Reflexes to Manage Generative Visibility

Several reflexes are suggested for communication departments. First, audit the presence in ChatGPT, Gemini, and Perplexity, focusing on the questions actually posed by prospects. Next, identify sources of influence by mapping the media and journalists that AIs cite in the sector, even if it means going beyond an established press list. For content, it is recommended to opt for formats presenting well-structured and referenced facts, such as press releases, proprietary datasets, or expert analyses. Ensuring coherence is essential: a message relayed multiple times in recognized media will have more impact than a one-off intervention. Finally, establish continuous monitoring and consider AI visibility as a standalone indicator, on par with outcomes. These principles are at the core of media relations: building narratives, placing messages, obtaining credible mentions, and multiplying citations across relevant channels.

Challenges for Brands: Risks, Shift to AI, and Expected Benefits

AI responses influence brand perception and purchase intent from the very first interaction. Without action, a brand can be poorly described or absent, losing valuable visibility, while incorrect or outdated information can be disseminated on a large scale. Visibility is no longer just about a results page: it is also built into generated responses, in a context described as having evolved since 2026. To exist where users ask their questions, the strategy must connect SEO, press relations, and data, with the awareness that systems describe and recommend based on sources that often escape direct control. It is possible to strengthen one’s presence in these spaces, and Cyndie Bettant summarizes the challenge by emphasizing the importance of being cited. Historically focused on Google and its levers, marketing teams have benchmarks to evolve toward this new framework. Brands that engage in a dedicated PR strategy for AI platforms are seen as likely to achieve better visibility and more complete and structured responses.