October 22, 2026 By the Editorial Desk (Adapted from MarTech reporting)
Main Facts: The Reality Check on AI Search vs. Traditional Organic Traffic
For the past two years, digital marketers have been swept up in a gold rush. The emergence of generative AI search engines, Large Language Models (LLMs), and chatbot interfaces birthed an entirely new discipline: Generative Engine Optimization (GEO). Agencies promised that ranking in ChatGPT answers or Perplexity summaries would be the new frontier of digital visibility, urging brands to reallocate massive budgets away from traditional SEO.
However, a sweeping new industry analysis suggests that the panic—and the hype—might be severely overblown.
According to a comprehensive study conducted by digital marketing veteran Marie Haynes, which analyzed 12 months of Google Analytics 4 (GA4) data across 100 websites and more than 145 million sessions, ChatGPT accounts for a microscopic fraction of web traffic. Specifically, ChatGPT sent just 0.62% as many sessions as Google organic search. Even when combining referrals from all other major LLMs into a single cohort, the total figure rises only marginally to 0.73% of Google organic traffic—not total website traffic, but merely a fraction of what Google alone provides.
Despite this staggering disparity in volume, the study uncovered a fascinating nuance: users arriving via ChatGPT convert at a significantly higher rate. Among the subset of 55 websites in the study actively tracking conversions, 13.8% of ChatGPT visitors completed a desired action, compared to just 3.02% of Google organic visitors.
Yet, volume matters. Because Google’s sheer traffic scale dwarfs that of LLMs, Google organic search ultimately drove 27 times as many total conversions as all LLMs combined.
As Haynes bluntly summarized in her findings:
"The AI search that is actually sending people to businesses today isn’t ChatGPT. It’s Google."
Chronology: How the AI Referral Wave Rose and Stalled
To understand where GEO stands today, it is helpful to look back at the trajectory of AI-driven referrals over the past few years.
Late 2023 – Early 2024: Following the public explosion of ChatGPT and the mainstreaming of conversational AI, early adopters noticed erratic but growing spikes in referral traffic coming from OpenAI’s domain. Digital marketing forums buzzed with theories on how to reverse-engineer OpenAI’s crawler behavior.
Mid 2024 – Mid 2025: The industry saw exponential growth in AI referrals. As OpenAI rolled out web-browsing capabilities natively within ChatGPT, users increasingly relied on the chatbot not just to write code or draft emails, but as an alternative search engine for product recommendations, recipes, and complex research. Agencies rushed to market with "GEO audits" and "AI visibility scores."
August 2025: According to Haynes’ data analysis, this exponential growth hit a hard ceiling. In August 2025, ChatGPT referral traffic plateaued across the board.
Late 2025 – Late 2026: Following the peak in mid-2025, referral traffic from ChatGPT actually began to experience a noticeable decline. Users increasingly consumed information directly inside the AI interface without clicking through to publisher websites, reducing outbound referral loops. This culminated in the October 2026 release of Haynes’ study, which formally categorized GEO as a "distraction" relative to its actual impact on enterprise bottom lines.
Supporting Data: Breaking Down the Numbers
To appreciate the scale of the gap between traditional search and generative AI referrals, we must look closer at the metrics provided in the multi-site analysis.
Metric Measured
ChatGPT / LLMs
Google Organic Search
The Disparity
Relative Session Volume
0.62% (ChatGPT) / 0.73% (All LLMs)
100% (Baseline benchmark)
Google drives roughly 135x to 160x more raw traffic.
Conversion Rate (Subset of 55 sites)
13.8%
3.02%
ChatGPT visitors are roughly 4.5x more likely to convert.
Total Conversions Driven
1 share
27 shares
Google organic produces 27x more total conversions overall.
Limitations of the Study
While the dataset—spanning 145 million sessions over 12 months—is robust, industry analysts have noted a few limitations to keep in mind:
The exact reporting dates and specific industry breakdowns (e.g., e-commerce vs. B2B SaaS vs. publishing) were not explicitly detailed in the public breakdown.
Definitions of "conversions" likely varied across the 55 participating sites, ranging from newsletter sign-ups to high-ticket product purchases.
The study did not measure year-over-year fluctuations in Google organic traffic itself, leaving open the question of how much traffic Google has lost internally to its own AI Overviews.
Official Responses and Expert Reactions: Is GEO Really a "Distraction"?
Marie Haynes’ bold assertion that GEO is a "distraction" has ignited fierce debate across the SEO and digital marketing communities.
On one side of the aisle, traditional SEO purists feel validated. For months, agencies have upsold clients on expensive AI optimization strategies that target LLMs, diverting resources away from core technical SEO, content quality, and user experience on traditional search engines. Critics argue that spending thousands of dollars trying to optimize for a channel that yields less than 1% of Google’s traffic volume is a poor allocation of capital.
On the other hand, forward-thinking digital strategists warn against writing off AI search entirely. They point out a critical flaw in viewing AI search only through the lens of direct referral traffic: intent transformation.
As highlighted in recent MarTech analyses, AI search engines are fundamentally transforming how consumers conduct pre-purchase research. When a user inputs a long, highly personal, multi-layered prompt into an LLM ("Compare enterprise project management tools that integrate with Salesforce and have native time-tracking under $20/user, and explain the main drawbacks of each"), they are not just looking for a link—they are conducting deep customer research. Even if the user doesn’t click a direct referral link immediately, being cited as an authoritative source in that AI-generated response builds brand equity and shapes consideration sets.
Furthermore, Google itself has integrated generative AI heavily into its own ecosystem via AI Overviews. While Haynes’ study specifically measured outbound referrals from standalone chat interfaces like ChatGPT, the lines between traditional Google organic search and generative search continue to blur.
Implications for Marketers: Where Should You Focus Your Budget?
As the dust settles on the initial AI hype cycle, digital marketers must adopt a pragmatic, balanced approach to search strategy. Here are key takeaways for navigating the search landscape:
1. Don’t Abandon Core SEO for GEO
The data proves unequivocally that Google organic search remains the undisputed king of digital acquisition. For the vast majority of businesses, a dollar spent improving site speed, fixing technical SEO errors, earning high-authority backlinks, and creating genuinely helpful human-first content will yield exponentially higher returns than chasing algorithmic quirks in ChatGPT.
2. Recognize High-Intent AI Traffic
If your analytics do show traffic trickling in from ChatGPT, Perplexity, or Claude, treat those visitors like gold. The data shows they convert at nearly 14%, meaning they arrive with exceptionally high intent and readiness to buy. Ensure your landing pages are tailored to capitalize on these highly qualified, low-volume visitors.
3. Monitor Brand Mentions, Not Just Clicks
AI visibility isn’t just about direct referral links anymore. Consumers are using LLMs to research brands, read synthesized reviews, and compare options. Even if an AI doesn’t link back to your site, appearing as a recommended authority in AI responses influences downstream branded search volume on Google. Marketers need to track both source citations and brand mentions in AI ecosystems.
4. Keep Pace with Google’s Evolving Guidance
Google continues to update its content guidelines—recently emphasizing the need for authentic human review in content creation—while refusing to guarantee that compliance will automatically reverse traffic drops caused by AI Overviews. Marketers must remain agile, focusing on creating deep subject-matter expertise that satisfies both traditional algorithms and AI synthesis engines.
Final Verdict
GEO is not dead, but it is deeply misunderstood. It is not currently a replacement for traditional search engine traffic, nor is it a magic bullet for mass customer acquisition. Marketers who keep their priorities straight—prioritizing the massive, reliable scale of Google organic search while quietly positioning their brands to be trusted authorities within AI ecosystems—will be best positioned to win in the years ahead.
