What actually drives citations in generative search engines
Source published ·Modelwire updated
Original coverage: arXiv cs.CL ↗·How Modelwire adds context
The development
A large-scale empirical study reveals how generative search engines like ChatGPT and Gemini select sources for citations, challenging the emerging GEO industry's assumptions. Analyzing nearly 160,000 answers across two markets, researchers found that content relevance (title and paragraph matching) drives citations within a single website, but structural elements like FAQs and page speed matter only across different sites. Critically, page-level factors explain minimal variance in citation patterns, suggesting model-internal knowledge and retrieval mechanisms play a larger role than publishers typically optimize for. This work reframes the SEO-to-GEO transition as fundamentally different, with implications for content strategy and the viability of generative engine optimization as a discipline.
Modelwire’s AI-generated summary of coverage from arXiv cs.CL.
Modelwire analysis
Analyst takeOur AI-generated reading of the wider context and the next developments to watch.
The study isolates which factors matter at which scale. Content relevance drives citations within a site, but page-level SEO signals (speed, structure) barely move the needle even across sites. The implication: publishers chasing traditional GEO tactics may be optimizing the wrong levers entirely.
This connects directly to the legal citation audit from October 8th, which found that models cite sources without necessarily grounding decisions in them. Here we see the inverse problem: models do ground answers in content, but the citation selection process is opaque to publishers. The Wikipedia traffic decline story (October 13th) adds urgency: if citation patterns are driven by model internals rather than page signals, then traditional content strategies fail to capture traffic from generative interfaces. Together, these three pieces suggest the SEO-to-GEO transition isn't just different in degree but in kind. Publishers are building for a system that doesn't respond to their inputs the way search engines do.
If Perplexity or other generative search engines publish their own citation methodology papers in the next six months, watch whether they confirm or contradict these findings. If they remain silent while GEO agencies continue charging for optimization services, that signals the industry is knowingly selling solutions to a problem the vendors haven't validated.
This interpretation is generated from the summary above and the archive coverage cited below. Our methodology · Report an error
Coverage behind this analysis
These archive entries ground the connection in our analysis. They are ordered by source publication date, with links to our coverage and the original sources.
·arXiv cs.CL
Legal LLMs cite the right laws but don't follow them
A new audit reveals a critical gap in how large language models justify legal reasoning. Researchers tested whether models actually ground their verdicts in the authorities they cite, by swapping named statutes and precedents for unrelated ones while holding facts constant. Results show models reliably name the correct legal source 67-100% of the time, yet…
MentionsChatGPT · Gemini · Perplexity · arXiv
How this coverage is produced
Modelwire uses AI to generate summaries and context from source headlines, snippets, and selected archive coverage. Automated checks do not verify every claim, and items are not routinely reviewed by a person before publication. Zacaria Solis operates the site. Read the linked source for the full evidence and report errors through our corrections process.
Modelwire summarizes, we don’t republish. arXiv cs.CL originally reported this story as “What Drives AI Citations? Content, Retrieval and Model Knowledge in Generative Engine Optimization (GEO)”. The full content lives on arxiv.org. If you’re a publisher and want a different summarization policy for your work, see our takedown page.