Research indicates that Google AI's summaries are influenced by Facebook.

      When Google’s AI responds to questions, it is increasingly relying on social media content rather than directly reading websites. Recent data indicates that social platforms have become a significant source for Google’s AI responses, often without the user ever visiting those platforms.

      These insights come from BrightEdge, an enterprise SEO company that monitors around 300 million searches each month. While the figures are proprietary to BrightEdge, the trend they reveal is difficult to overlook.

      The statistics

      According to BrightEdge, Facebook was cited as a source in 19.5 million Google AI Overviews, while Instagram showed up 877,000 times, and TikTok 78,000 times. In other terms, the firm estimates that one in every 15 searches is now addressed using social media. This situation positions Meta’s platforms as the primary contributors to Google’s AI responses.

      This change reflects the operational nature of the AI. Google’s AI acts more like a researcher, drawing on broader discussions related to a topic instead of directing users to a specific brand's website. Posts, videos, community discussions, and creator content all play a role in shaping the answers generated.

      Different platforms as distinct authorities

      A notable assertion is that Google’s AI does not perceive social media as a monolithic entity. It seems to assign different platforms specific roles based on their strengths. BrightEdge describes this phenomenon as an expert network: Facebook is relied upon for local and community insights, Instagram is associated with lifestyle and culture, TikTok is known for viral trends and deals, Reddit is valued for firsthand experiences, and YouTube offers how-to guidance.

      Concrete examples illustrate this point: Facebook appeared for urgent, practical inquiries like car recalls or the operational status of a theme park. Meanwhile, Instagram was highlighted for culture and shopping-related queries, such as a celebrity’s engagement ring or ferry options to the Amalfi Coast.

      The study even references specific posts, noting that a Facebook post from the Cincinnati Reds influenced a live sports-related query that garnered an estimated 11.6 million monthly searches. Additionally, an Instagram post was linked to searches about mobile payment applications.

      Implications for brands

      For marketers, this development poses a challenging message. Gaining visibility in AI responses is no longer synonymous with simply ranking a website. Jim Yu, CEO of BrightEdge, emphasized the importance of this shift, pointing out that Google’s AI has a social dimension which alters how brands must approach visibility. He argued that a single post, video, or discussion can significantly affect the answer presented to a customer.

      This concept is known as generative engine optimization, where the aim is to shape the AI's responses rather than just a webpage's ranking. This aligns with the current effort to secure citations in AI answers, amid a broader debate regarding compensation when AI utilizes content from the open web.

      There is also a potential drawback for brands: influence might arise from accounts they do not directly manage, including smaller ones. A specific niche Reddit discussion or a local group could have more sway than a well-crafted corporate page.

      The broader context

      This data emerges at a time when Google’s AI-generated answers are under increasing scrutiny for their accuracy and sourcing. Instances have been noted where these overviews mistakenly presented fictional information as fact, making the choice of sources a critical issue.

      If social media posts are indeed shaping answers for millions of queries, then the quality of those posts becomes essential. Additionally, platforms owned by Meta are increasingly acting as an unnoticed layer within Google’s search framework, providing substantial data directly to a competing service.

      At present, the key takeaway is straightforward: social media is no longer merely a platform for scrolling; it is becoming a primary resource for Google’s AI in its quest for information.