SimilarWeb figures reveal a large audience overlap and a vast traffic gap. SEO experts say the work of search optimization will change rather than vanish.
Data compiled from August show that almost every person who used ChatGPT also visited Google in the same month. SimilarWeb numbers shared by SEO analyst Brodie Clark[1] and Aleyda Solis put the overlap at 95.3 percent, while the reverse overlap stood at 14.3 percent. In raw visits, ChatGPT drew about 5.8 billion sessions that month,[2] compared with roughly 83.8 billion for Google.
What the numbers mean
The figures underline a simple reality. Generative AI tools are growing fast, but many users still turn first to established search engines. Separate tracking from Datos and SparkToro finds traditional search engines still account for roughly 95 percent of search activity. User behavior studies also show heavy users of AI tools continue to verify or supplement results with conventional search engines.
Traffic patterns to publisher sites reflect mixed outcomes. Some industry reports claim referrals from ChatGPT have cooled and that Bing now sends more visitors to certain sites. At the same time, Search Engine Land reported a sharp rise in referrals from ChatGPT in August, showing a 1,233.5 percent increase and outpacing Bing for their site. Even after that spike, Search Engine Land recorded 37 times more referral traffic from Google organic search than from ChatGPT.
Revenue and relevance
Early indicators suggest LLM-driven referral traffic still translates into relatively little direct revenue for many publishers. Survey respondents in the SEOFOMO AI Search Optimization study said generative AI channels deliver under 5 percent of revenue for most sites, while traditional organic search remains responsible for more than half of site revenue on average. That revenue gap keeps search engine optimization central to many publishers’ business models.
LLMs also depend on up-to-date external sources when current facts matter. When models need fresh information they rely on search integrations and live retrieval. Independent testing by some SEOs and reporting by media outlets has shown that ChatGPT can draw on results from multiple search providers, including Google, when it grounds responses.
What stays the same, what changes
Practices that underpin discoverability remain relevant. Principles such as crawlability, indexability, content relevance, signals of expertise and trust, and popularity among target audiences still matter when content must be found and cited. What shifts are the platform-specific optimization details. Engineers and content teams must account for different user agents, reduced capacity for client-side rendering on some platforms, changing query patterns, and a heavier emphasis on verifiable citations inside AI answers.
SEO professionals note that the discipline has never been only about one interface. Search optimization adapted when YouTube and then TikTok became major discovery channels. The same adaptation will apply to AI-powered surfaces. Optimization will aim not only at a traditional search results page but at the places and APIs where models retrieve and present answers, whether that is in a chatbot, a copiloting feature inside an app, or voice assistants.
Industry view and likely winners and losers
Voices across the industry push back on any claim that SEO is dead. Many practitioners frame the moment as a mutation. They argue that some kinds of publishers will feel the pain sooner. Sites built around quick informational clicks such as how-to articles, recipe lists, trivia, and affiliate roundups may lose volume as models surface full answers. Smaller newsrooms and niche publishers without strong brand recognition may also see a drop in direct visits.
By contrast, local services, sites with established trust in health and finance verticals, and businesses that deliver transactive or high-trust interactions are likely to retain stronger positions. For those properties AI tools tend to act as assistants that complement rather than replace discovery and commerce.
A separate strand of criticism focuses on how the search landscape evolved before AI. Some observers say search has long shifted toward paid placements and blended result formats that reduce organic visibility. That history shapes how publishers approach the next phase of optimization.
Practical takeaway
The near-term picture is straightforward. Large numbers of users now engage with both generative AI and classic search engines. Publishers should treat AI as an additional distribution channel and not as a replacement. Tracking referral sources closely, testing whether AI-driven traffic converts, and adapting technical SEO to the retrieval needs of LLMs are sensible priorities. Over time the objective will be to surface brand and content inside an expanding ecosystem that includes Google’s AI features, ChatGPT and other LLM interfaces, Perplexity, copilot-style assistants, and voice platforms.
In short, the data point to rapid growth for generative tools. They do not show that established search has been displaced. The task for publishers and SEO teams is to evolve tactics so a site can succeed across both legacy and emerging discovery systems.
Notes: This post was edited/created using GenAI tools.
Read next: TikTok Revenue Surges, ChatGPT Leads Installs, Streaming Platforms Expand Earnings as August App Rankings Stabilize[3]
References
- ^ SEO analyst Brodie Clark (x.com)
- ^ 5.8 billion sessions that month, (www.digitalinformationworld.com)
- ^ TikTok Revenue Surges, ChatGPT Leads Installs, Streaming Platforms Expand Earnings as August App Rankings Stabilize (www.digitalinformationworld.com)