A 45-Study Survey Keeps Foundational GEO Gains Inside a Fixed Context
The 15 July 2026 arXiv survey of 45 studies says the famous visibility gains assume the source is already in context.
Direct answer
Olivier Martinez's 15 July 2026 survey reviews 45 studies from November 2023 through July 2026 and says foundational GEO gains are conditional on a source already sitting in a fixed context. No reviewed technique shows a stable cross-platform effect on organic discoverability.
Optimizing Visibility in Generative Engines: A Critical Survey of Generative Engine Optimization (2023-2026), arXiv:2607.14035, is dated 15 July 2026 and credited to Olivier Martinez. The version note says the survey covers work made public or formally accepted from the first version of the foundational paper, released on 16 November 2023, through 14 July 2026. The abstract says it reviews 45 studies in that window, including one earlier preprint published at EMNLP after the window opened, plus relevant RAG and evaluation work.
The abstract argues that GEO is not a single ranking task. It describes a stochastic, partially observable pipeline spanning search activation, crawling and indexing, retrieval, reranking and context allocation, citation, prominence, factual absorption, fidelity, and user behavior. It says the foundational paper's widely cited gains are valid inside their experimental setting but conditional on a source already being present in a fixed context. Those gains establish neither organic discoverability nor durable traffic effects.
Reviewed work, the abstract says, indicates that topical relevance and context position are the most reproducible levers, that generic heuristics transfer poorly, that competition can erode individual gains, and that citation-oriented rewrites can impair retrieval. Commercial audits in the same paragraph show low source overlap, substantial run-to-run variability, and persistent fidelity gaps. The contribution list is a multistage formal model, a visibility vector that separates discoverability, citation, absorption, and economic outcomes, an evidence hierarchy, and a protocol based on repeated measurements, paraphrases, controls, human validation, and multi-actor interference.
The narrow conclusion
The abstract's last sentence says the evidence in this corpus is narrow: already-retrieved content can causally alter its citation or use, but no reviewed technique shows a stable, longitudinal, cross-platform causal effect on organic discoverability or downstream behavior. The introduction adds that the up to 40 percent figure is often recast as a general promise, although it describes a relative visibility gain in a simulator where five documents are already in context.
Source: Martinez, arXiv:2607.14035.
FAQ
- How many studies are in the review window?
- The abstract says 45 studies, plus relevant RAG and evaluation work, from November 2023 through July 2026.
- Which levers does the abstract call most reproducible?
- Topical relevance and context position.