For years, small business owners have measured online success by the same yardstick: Google traffic. Rankings, clicks, organic sessions — these numbers defined whether your website was working. A study published in July 2026 quietly rewrites that assumption, and the implications for any business that sells something online are significant.
Previsible, a search analytics firm, released its third annual State of AI Discovery Report after analyzing 6.77 million LLM-driven sessions across 166 websites spanning industries from e-commerce to legal services. The headline finding: ChatGPT traffic converts at roughly nine times the rate of standard Google organic traffic.
What the Data Actually Shows
The study tracked sessions sourced from the major AI platforms and compared conversion behavior against baseline organic traffic. The conversion rate spread was striking:
- ChatGPT: ~15.9% conversion rate
- Perplexity: ~10.5%
- Claude: ~5%
- Gemini: ~3%
- Google organic (baseline): ~1.76%
The researchers also found that AI referral visitors spend nearly twice as long on site and view significantly more pages per session than typical organic visitors. This is not a rounding error — it reflects a structural difference in how people use AI search versus traditional search.
Why AI Visitors Buy More
The most plausible explanation is intent compression. When someone types a question into Google, they are often early in their journey: they might browse five results, compare options, then disappear. When someone asks ChatGPT or Perplexity a specific question and the AI recommends your business by name, that person has already done significant research inside the AI conversation. They arrive at your site pre-convinced, looking to confirm rather than compare.
This is why the volume being small right now does not undermine the argument. Previsible found that monthly LLM-sourced sessions grew 9.9x in 19 months, reaching 644,478 sessions in May 2026 across their study sample. The channel is tiny today but growing fast — and the quality of the traffic is already high.
The Measurement Gap Most Businesses Have Not Solved
Here is the problem: most small business owners cannot see this traffic at all. Standard Google Analytics reports lump AI referrals from ChatGPT, Perplexity, and Claude into the generic referral bucket, often mixed with other low-quality referral sources. The signal is buried.
Google took a step toward fixing at least part of this picture in June 2026, when it launched Generative AI Performance Reports directly inside Search Console. The new reports, available under the Performance tab, show impressions from AI Overviews and AI Mode separately from standard search results — giving website owners, for the first time, a dedicated view of how their content appears inside Google's AI-powered features. (Google Search Central, June 2026)
The current version shows impressions, pages, countries, and devices — but not clicks yet. Even so, if your pages are showing zero impressions in the AI tab, that tells you something important: your content is not being surfaced inside Google's AI features, regardless of your traditional ranking position.
The 28.8% Problem You Did Not Know You Had
The Previsible study uncovered one more finding that has direct operational implications: ChatGPT sends 28.8% of its referral traffic to a site's internal search results page, rather than to the specific content page the user was asking about. The AI trusts your domain enough to recommend it, but it cannot always identify the right URL — so it defaults to your site's own search box.
This means that even when you are being cited by ChatGPT, nearly a third of those visitors land on a dead-end search results page that may return nothing useful. If your site search is broken, empty, or nonexistent, you are losing a meaningful share of your highest-converting traffic before they see your services.
Four Things to Do Right Now
- Audit your site search. Go to ChatGPT and ask it about services you offer in your city. If it links to your site, click the link — does it land on the right page? If it lands on a search results page, test what your internal search returns. Fix gaps before spending any money on outreach.
- Connect Google Search Console and check the AI tab. Under Performance → Generative AI, look at which pages earn impressions in AI Overviews. Pages with zero impressions are invisible to Google's AI layer, no matter their organic ranking.
- Set up AI referral tracking in GA4. Create a custom segment filtering sessions where the source contains chat.openai.com, perplexity.ai, claude.ai, and gemini.google.com. This separates high-converting AI sessions from generic referral noise.
- Prioritize third-party citations. AI engines cite sources they trust. Independent directory listings (Clutch, Yelp, Google Business Profile), real customer reviews, and mentions in local publications are the input variables that determine whether AI recommends you in the first place.
The Bilingual Opportunity
For businesses serving both English and Chinese-speaking customers in New York — and for the agencies building their digital presence — the AI search gap has a specific dimension. Most local businesses do not have content in both languages that AI engines can confidently cite. A Chinese-language page that appears authoritative, is properly indexed, and is referenced by credible third-party sources in Chinese is a largely uncontested position in most local markets. The volume of Chinese-language queries entering AI tools is real and growing, and the supply of credible, citeable local business content to answer them remains thin.
This is the window. The businesses that establish third-party credibility now — through directory listings, reviews, and cited content — will be the ones AI engines recommend as the channel matures.