AI Search Ignores Syndicated News and Press Releases
▼ Summary
– A BuzzStream report analyzing AI citations found that press releases distributed through syndication channels (like Yahoo or MSN) account for only 0.04% of all citations, challenging services that market them for AI visibility.
– Original editorial content from news publications dominated AI citations, making up 81% of news citations, while affiliate and review content accounted for the rest.
– One key exception is that ChatGPT frequently cites internal press releases from company-owned domains (18% of its citations), unlike Google’s AI platforms where such citations are around 3%.
– The data suggests that earned editorial coverage is far more likely to be cited by AI than distributed press releases, aligning with observations that being mentioned by other sites aids AI recommendations.
– The report’s findings have limits, including a one-week data window and a focus on large brands, meaning results for smaller brands with less existing coverage may differ.
A recent analysis of millions of AI-generated answers reveals a significant gap between marketing claims and reality regarding press release visibility. While many distribution services promote their ability to get content featured in AI search results, the data tells a different story. Original editorial content from reputable news outlets dominates the citations, leaving syndicated press releases and news with a negligible presence.
The comprehensive study examined over four million citations across major AI platforms, including ChatGPT and Google’s various AI search modes. Researchers used thousands of prompts across multiple industries to gather data. The findings were stark: while news publications accounted for a notable portion of all citations, the specific breakdown was telling. Press releases distributed through major syndication channels like Yahoo and MSN represented a minuscule fraction of the news citations. Direct citations from prominent newswire services themselves were almost as rare.
In contrast, the research highlighted what does gain traction with AI systems. Original reporting and editorial content constituted the overwhelming majority of news citations. This type of content appeared consistently, whether users were asking evaluative questions, seeking brand information, or looking for general facts. Affiliate and review content filled the remainder of the news category, performing particularly well in prompts where users compare products or services.
One intriguing platform-specific finding emerged from the data. ChatGPT showed a distinct tendency to cite internal press releases and newsroom content published directly on a company’s own website. This pattern was far less pronounced on Google’s AI platforms. For instance, when asked about specific corporations, ChatGPT frequently sourced answers from the companies’ official press rooms, even referencing older announcements.
This research directly challenges a growing marketing narrative within the public relations industry. Several press release distribution services have begun actively promoting AI search visibility as a key benefit of their syndication networks. The data suggests, however, that the version of a press release that gets blasted across hundreds of syndicated sites is rarely the one an AI model chooses to reference. The version that earns genuine media pickup or resides on an owned corporate domain holds far more weight.
This insight connects to broader trends in how AI systems gather and prioritize information. Many top publishers now restrict access to their sites for AI training bots, which could influence what content is available for models to learn from and later cite. Furthermore, the way AI models are designed to answer questions may inherently favor certain content types. As one Google executive noted, these systems often operate like a person conducting research, seeking out trustworthy and context-rich sources. Earned media coverage in established publications naturally carries more authority in this context than a distributed press release.
The implications for communication and marketing strategies are clear. For brands seeking visibility in AI-powered search and answer engines, the focus should likely shift. Investing in relationships that lead to genuine editorial coverage may be more effective than relying on broad press release distribution alone. The channel through which information is disseminated appears to be a critical factor in whether AI systems deem it a citable source.
It is important to note the study’s scope, which focused on data from a single week and included many large, well-known brands. The landscape for smaller entities with less existing media coverage could differ. This report is the first part of a larger analysis, indicating that our understanding of AI citation patterns will continue to evolve. For now, the evidence strongly suggests that in the world of AI search, quality and origin of content significantly outweigh the breadth of its distribution.
(Source: Search Engine Journal)
