ChatGPT is becoming a more significant discovery surface in the United States. Semrush's Traffic Analytics data for July 2026 estimates that chatgpt.com received about 1.09 billion visits, a 48.38% increase year over year. In the same dataset, bing.com referrals declined about 50.43%, while Google's direct figures rose 10.72%. The numbers do not mean traditional search has been displaced, but they do show that AI-assisted discovery is growing faster than established channels.
For marketers and website owners, the more immediate issue is measurement. Traffic that begins in an AI chat interface can be harder to identify and attribute consistently than a conventional search visit. That makes it risky to judge channel performance using a single analytics view, particularly when businesses are trying to understand whether their content is being discovered through ChatGPT, Google, Bing, or a combination of those paths. The Semrush US trending websites data is based on Traffic Analytics clickstream estimates across desktop and mobile. Semrush says its panel includes roughly 200 million anonymized users. These figures are therefore third-party estimates, not first-party traffic reports from OpenAI, Google, or Microsoft, but they provide a useful directional view of changing traffic patterns.
| Surface or metric | July 2026 US signal | Year-over-year change |
|---|---|---|
| chatgpt.com visits | Approximately 1.09 billion visits | Up 48.38% |
| bing.com referrals | Referral activity tracked in the same dataset | Down about 50.43% |
| Google direct figures | Google remained the largest referrer | Up about 10.72% |
The table should not be read as a like-for-like market-share comparison. It combines different signals reported within the same analysis, including visits and referrals. Its value is in the contrast: ChatGPT's growth was substantially faster than Google's, while Bing showed a sizable decline in referrals.
Semrush's wider research adds context. Its April 2026 study reported that outbound referral traffic from ChatGPT grew 206% during 2025. It also found that Google represented a sizable share of ChatGPT referrals. In July, Semrush reported that roughly 21.6% of ChatGPT referrals went to Google, and more than 30% of referrals went to the top 10 domains. That concentration matters because growing AI referral traffic may not be distributed evenly across the web.
AI-assisted discovery is not necessarily a replacement for search engines. A user may begin with a conversational question, receive a response with links or search prompts, then continue their journey on Google or another destination. Semrush's referral findings suggest that ChatGPT can also send meaningful traffic toward Google.
For businesses, this means the practical question is not simply whether to choose AI platforms or search engines. It is whether useful, trustworthy content can be found and cited across both environments. A clear service page, product explanation, help article, or original research asset may support conventional search visibility while also giving AI systems more material to surface or reference. GA4 remains useful for measuring owned web activity, but AI-driven referrals create attribution challenges. Referral data depends on how the originating surface passes traffic to a browser and how that visit is captured. In AI chat interfaces, those pathways may not always appear as a clean, consistent referral source.
That does not make GA4 unusable. It does mean teams should avoid treating a low or missing AI referral count as proof that AI discovery is irrelevant. The underlying visitor may arrive through a link, a copied URL, a follow-up search, or another path that obscures the original interaction.
A more resilient reporting approach is to combine several signals:
The July data does not justify abandoning Google-focused SEO. Google remained the largest referrer in Semrush's analysis, and its reported growth was still positive. Instead, it supports a broader content and measurement strategy that does not assume every valuable discovery journey starts with a conventional search result.
Prioritize pages that answer specific customer questions directly, identify the business and its offerings clearly, and provide details that can be verified on the page. Keep important information accessible instead of burying it in image-only assets or vague marketing copy. Then measure whether these pages generate engaged visits, inquiries, and revenue, not merely rankings or referral labels.
AI discovery is becoming another visibility layer that businesses need to understand. Scalevise helps teams assess where their brand appears in AI-generated answers, identify content gaps, and turn those findings into practical website improvements that support qualified discovery. Use the Scalevise AI Visibility and GEO Checker to see how your business shows up across AI search experiences and start an AI visibility scan.
How much did ChatGPT traffic grow in the US?
Semrush estimated that chatgpt.com received approximately 1.09 billion US visits in July 2026, up 48.38% year over year.
Did Google lose its position as the largest referrer?
No. Semrush's analysis indicated that Google remained the largest referrer. Its direct figures rose about 10.72% year over year, although that was slower than ChatGPT's reported growth.
Does the Semrush data prove that ChatGPT is replacing search engines?
No. The data shows a strong increase in ChatGPT traffic and changing referral patterns. Semrush also reported that Google receives a sizable share of ChatGPT referrals, indicating that AI-assisted discovery and search can be connected.
Why can AI referral traffic be difficult to measure in GA4?
AI chat surfaces can pass visitors to websites through pathways that are not always captured as a consistent referral source. As a result, analytics reports may not fully show the role AI played in a visitor's discovery journey.
Semrush's July 2026 data offers a clear signal that ChatGPT is growing rapidly as a US discovery surface, even as Google remains the dominant referrer. The priority for businesses is not to declare traditional SEO obsolete. It is to build content that can serve both search and AI-assisted journeys, while using measurement methods that account for incomplete attribution.