Ranking in Google won't get you cited in ChatGPT
A new July study shows top-10 rankings drive citations in Google AI and Perplexity but barely move ChatGPT. Here is how to work all three engines.
Ranking on Google's first page now earns you citations in two AI engines and skips the third. A July 2026 study from CiteLens found that 93% of Google AI Mode citations and 89% of Perplexity citations came from a query's top-10 organic results. In ChatGPT, only 30% did. So if you have been treating AI search as one channel to optimize, split it. Getting cited in ChatGPT takes different work than getting into Google's AI answers, and the gap is now wide enough to plan around.
What the CiteLens study measured
CiteLens ran 320 templated buyer queries across three consumer sectors in June 2026, then checked which sources each engine cited. The pattern was sharp. Google AI Mode showed a 0.92 correlation between a citation and a top-10 ranking. Perplexity showed 0.87. ChatGPT showed close to zero. Seventy percent of ChatGPT's cited sources ranked outside the top 10 on both Google and Bing, and fewer than 4% sat in Bing's top 10.
Two caveats keep this honest. The sample was Turkish-market queries in three verticals, so read it as a strong signal and not a global law. And Claude landed in the middle, with about half its citations from top-10 results and 58% going to sites that have a Wikipedia page. The full write-up is here. The direction matches what practitioners have watched all year, so the takeaway holds even if your own numbers differ.
Why ChatGPT ignores your rankings
Google AI Mode and Perplexity read the live search index, so your ranking is their shortlist. ChatGPT works from a different supply: its training data, its own retrieval, and a growing stack of licensing deals. On 23 July, Yelp licensed 330 million reviews and more than 8 million business listings to OpenAI, with Yelp's "Request a Quote" coming to ChatGPT next. Yelp's stock jumped 8% on the news. Picture someone asking ChatGPT for the best plumber in Denver. The answer now leans on Yelp's licensed data, not on whoever ranks first for "Denver plumber". Ranking did not put that business in the answer. A strong review profile and a licensing pipe did.
The deal is non-exclusive, so expect more of them. Every data partner OpenAI signs pulls another slice of ChatGPT's answers further away from the open web you can rank in.
For SaaS, the review sites are your Yelp
The Yelp story is local and consumer, but the mechanism is the same for B2B software. When a buyer asks ChatGPT to compare tools or name the best option for a use case, the model leans on G2, Capterra, Reddit threads, and the roundup posts that already list you. Your own landing page, however well it ranks, is one voice among those. So keep your G2 and Capterra profiles current and well reviewed, get into the honest "best X software" roundups your buyers read, and give real answers in the communities where they compare notes. That is the presence ChatGPT samples when it recommends software.
Work each engine on its own terms
For Google AI Mode and Perplexity, nothing changes. Rank in the top 10 for the queries that bring revenue, and you show up in their AI answers. Classic SEO is still the lever, and it now pays twice.
For ChatGPT, chase presence instead of position. Earn mentions on the third-party pages your buyers already trust. Build the entity signals these models rely on, starting with a clean, sourced Wikipedia page if you qualify for one. This is the same muscle behind branded search in the AI era: the more the web treats you as a known name, the more often the model names you back.
Measure citations by engine, not as a lump
A single AI-visibility score hides the split this study exposes. Track where you get cited engine by engine. Run your top buyer queries through Google AI Mode, Perplexity, and ChatGPT once a month, and log who each one names. Watch the referral side too, since ChatGPT already sends most of your AI referral traffic. When Google's AI cites you but ChatGPT never does, you can see it, and you know which lever to pull next.
Three engines, three sources, three plays. Keep ranking for Google and Perplexity. Build presence and reviews for ChatGPT. Measure all three apart, so your budget lands on the gap that is actually costing you visibility.
If you want help turning this into a per-engine plan for your own queries, that is the kind of work we do.
