AI found us before Google did | Hacker News

"Our content was structured with LLM readability as an explicit goal - not SEO. Consistent terminology, clear entity definition, a named methodology, and topical depth over breadth. LLMs seem to evaluate authority differently than search engines. Google proxies authority through external signals (links, engagement, domain age). LLMs appear to evaluate something closer to conceptual coherence - whether a source demonstrates genuine understanding of a subject in a way the model can parse and trust."

"There's a growing debate about whether "GEO" (Generative Engine Optimization) is a real discipline or just rebranded SEO. The SEO camp argues: good content + technical optimization = you appear in AI answers too. This case suggests the mechanisms are at least partially independent. We had zero SEO signals and nonzero AI visibility. You can apparently achieve one without the other."

"We weren't trying to rank. We were trying to be understood. And for at least one model, it worked. But it's made us think differently about what "visibility" means when the retrieval system is a language model rather than an index."

A startup offering AI visibility services received inbound leads from companies discovering them through Google Gemini despite having zero traditional SEO indicators: no Google Search Console setup, no backlinks, and no Google indexing. The site was structured for LLM readability with consistent terminology, clear entity definitions, and topical depth. This suggests LLMs evaluate authority differently than search engines. While Google relies on external signals like links and domain age, LLMs appear to assess conceptual coherence and genuine subject understanding. This case demonstrates that AI visibility and traditional search visibility may operate through partially independent mechanisms, challenging assumptions that good content automatically translates across both discovery systems.

Read at Ycombinator