The text discusses the complexities of how ChatGPT selects and cites its sources, which can be particularly relevant for marketing professionals. The analyses by Chris Green and Suganthan Mohanadasan reveal that ChatGPT’s source selection process is not transparent, as the final answers do not show how sources are chosen. This is important for marketers because it affects how their content might be retrieved and cited by AI, impacting visibility and reach.
The findings show that ChatGPT uses different retrieval pipelines, such as Labrador, Bright, Oxylabs, and SERP, which influence the sources selected. This variability means that marketers need to ensure their content is easily retrievable and readable by AI systems. For instance, content that is in plain HTML, has crawlable facts, and is text-heavy is more likely to be selected by ChatGPT.
Moreover, the research highlights that some queries are classified in a way that skips web searches, meaning that even if content is relevant, it might not be fetched or cited. This underscores the importance for marketers to optimize their content for AI readability and ensure it is accessible without complex scripts or hidden behind paywalls.
In summary, marketers should focus on creating content that is easily accessible and readable by AI systems like ChatGPT to improve the chances of their content being selected and cited. This involves optimizing for plain HTML, ensuring clear and crawlable information, and maintaining strong third-party coverage.
Kaynak: https://searchengineland.com/chatgpt-citations-change-hidden-search-pipelines-481843
