When Google’s AI answers a question, it is increasingly not reading a website. It is reading a Facebook post. New data suggests social platforms have become a core source for Google’s AI answers. Often the user never visits the platform at all.
The findings come from BrightEdge, an enterprise SEO firm. Its tool tracks roughly 300 million monthly searches. As with any single-vendor study, the numbers are BrightEdge’s own, but the direction of travel is hard to ignore.
The numbers
BrightEdge says Facebook appeared as a source in 19.5 million Google AI Overviews. Instagram appeared 877,000 times, and TikTok 78,000 times.
Put another way, the firm reckons one in 15 searches is now answered using social media. That makes Meta’s platforms, it says, the single biggest resource feeding Google’s AI answers.
The shift reflects how the AI works. Google’s AI behaves more like a researcher. Instead of pointing to a brand’s website, it pulls from the wider conversation around a topic. Posts, videos, community threads and creator content all feed the answer.
Every platform is a different ‘expert’
The most striking claim is that Google’s AI does not treat social media as one thing. It appears to trust different platforms for different jobs.
BrightEdge describes it as an expert network. Facebook covers local and community signals. Instagram handles lifestyle and culture, TikTok viral trends and deals, Reddit firsthand experience, and YouTube how-to context.
The examples make it concrete. Facebook surfaced for timely, practical queries, such as a car recall or whether a theme park was open that day. Instagram leaned to culture and shopping, from a celebrity’s engagement ring to the ferry to the Amalfi Coast.
The study even names specific posts. A Cincinnati Reds Facebook post fed a live-sports query estimated at 11.6 million monthly searches. An Instagram post led one on mobile payment apps.
What it means for brands
For marketers, the message is uncomfortable. Being visible in AI answers is no longer the same as ranking your own website.
Jim Yu, BrightEdge’s chief executive, put it plainly. “Google’s AI has a social side, and that changes how brands need to think about visibility.” A single post, video or thread, he argued, can shape the answer a customer sees.
This is generative engine optimisation. The goal is to influence what the AI says, not just where a page ranks. It sits alongside the scramble to earn citations in AI answers. There is also a wider fight over who gets paid when AI harvests the open web.
There is a catch for brands, too. Influence can come from accounts they do not control, including small ones. A niche Reddit thread or a local group can outweigh a polished corporate page.
The bigger picture
The data lands as Google’s AI answers face growing scrutiny over accuracy and sourcing. Overviews have been caught treating fiction as fact, so the choice of sources is more than an academic point.
If social posts shape answers for millions of searches, the quality of those posts matters. Meta’s platforms are also becoming a hidden layer of Google’s search. They feed rich data straight into a rival’s product.
For now, the takeaway is simple. The social feed is no longer just where people scroll. It is increasingly where Google’s AI goes to find the answer.