MarketingProfs’ July 24 AI Update put a review of 45 studies on generative engine optimization at the center of its weekly AI roundup, and the conclusion was blunt: the evidence does not support claims that GEO reliably boosts AI visibility. The review found no evaluated technique that consistently improved organic discoverability, downstream traffic, or business outcomes across AI platforms.
That matters because the best-known GEO promise in the market, the claim that visibility can rise by about 40%, now looks much thinner than the pitch deck version. The roundup says that figure traces back to a single lab study, not a broad or replicable body of evidence. A Seer Interactive article titled Optimizing Content for Generative Search Engines Resulted in 40% More Visibility helped spread that number, but the new review treats it as an isolated result, not a settled benchmark.
The skepticism lands against the backdrop of the field’s own origin story. The Princeton University paper GEO: Generative Engine Optimization, written by Pranjal Aggarwal, Vishvak Murahari, Tanmay Rajpurohit, Ashwin Kalyan, Karthik Narasimhan and Ameet Deshpande, first appeared on arXiv in November 2023 and was later accepted to KDD 2024. The ACM conference version was published on August 24, 2024, giving the term academic legitimacy even as commercial vendors and marketers rushed to turn it into a performance claim.
That gap between academic framing and market promises is now at the heart of the debate. A July 21 post on Stephen’s Lighthouse, citing MyClaw Newsletter, summarized the same 45-study survey as finding no GEO technique that reliably improves cross-platform discoverability, traffic or conversions. The message across those layers is consistent: the category is real, but the proof behind many of its boldest claims is not.
For marketers and publishers, the practical answer is to stop treating GEO as a guaranteed lever for AI search lift. Content structure, clarity and authority still matter, but the update argues against paying for promises that rest on weak proxies and one-off lab results. Budgets tied to AI visibility should be judged on measured citation exposure, traffic and business outcomes, not on a vendor’s preferred headline metric.
