Two AI-visibility vendors, Press Ranger and OtterlyAI, published a joint study on August 20 claiming that a publisher’s ChatGPT citation count rises 48 percent once it signs an OpenAI licensing agreement. The finding comes from the companies’ own commercial dashboard, not from a peer-reviewed or third-party source. That distinction matters because both companies sell products built on exactly the citation behavior the study describes.
OtterlyAI, a GEO monitoring platform, supplied 129.3 million citations pulled from seven answer engines during June 2026: Perplexity, Gemini, Claude, Microsoft Copilot, ChatGPT, and both of Google’s AI surfaces, AI Overviews and AI Mode. Press Ranger, a PR distribution company, matched that citation set against its own database of 91 confirmed AI licensing agreements spanning 314 publisher domains. The two companies then cross-referenced the datasets and released the results as promotional material for their respective platforms.
The headline figure, a 48 percent citation lift on ChatGPT for OpenAI-licensed publishers (10.2 citations per cited page versus 6.9 for unlicensed pages), held only on OpenAI’s own platform. Publishers who had struck deals with Google actually pulled a slightly lower citation rate on AI Overviews than unlicensed publishers with similar profiles, while Perplexity showed no gap at all between its licensed and unlicensed publisher sets. OpenAI stood alone among the three licensors with a measurable home-platform payoff, and that payoff more than doubled, to a 112 percent lift, among publishers who had licensed with OpenAI exclusively.
That platform-specific split is the study’s most useful signal for search teams weighing whether a licensing deal is worth pursuing. A deal with OpenAI appears, in this data, to concentrate citation gains on ChatGPT rather than spread them across the wider generative search field. Publishers chasing broad AI visibility across Google, Perplexity, and Copilot would not see a comparable return from an OpenAI-only agreement.
News itself is a minor slice of what these engines cite: 7.2 percent of the 129.3 million citations analyzed, per the study. Within that slice, concentration runs steep. Just five publishing groups (Forbes, Hearst, Conde Nast, People Inc., and Future plc) account for 69 percent of all citations earned by licensed outlets. Trade and niche publications, the large majority unlicensed, pulled in more total citations than mainstream outlets across 16 industries the study broke out, losing that lead in only one category, and outpacing mainstream media by 213 percent where they led.
Neither company disclosed an outside audit confirming the citation counts or the causal link between licensing status and citation volume. Press Ranger sells licensing research and press distribution; OtterlyAI sells citation tracking. Both stand to gain commercially if buyers treat licensing deals and citation monitoring as necessary spend, a structural reason to read the topline figure as directional rather than settled. No academic body or independent research group has replicated the citation counts.
For search and communications teams, the more durable takeaway sits in the study’s secondary finding. Roughly 47 percent of the citations earned by licensed publishers went to commercial evergreen formats, a category the study defines as product reviews, buying guides, and ranked recommendation posts, which suggests format drives citation volume at least as much as licensing status does. Teams weighing an AI licensing deal should treat it as a ChatGPT-specific bet rather than a general AI-visibility strategy, and should prioritize earning coverage in the trade and niche publishers that already dominate citations in their industry before assuming a licensing agreement is the faster route.
Press Ranger and OtterlyAI announced the findings in a joint press release distributed via GlobeNewswire on August 20, 2026.
