Landmark Post

Kleinberg's paper introducing the hubs-and-authorities model of link analysis, developed contemporaneously with PageRank and still the reference point for topical authority arguments.

The model Google deployed in Search in 2019 to better interpret query context and prepositions. The technical basis for the shift away from keyword matching toward passage-level meaning.

The paper that named generative engine optimization, proposing a benchmark and testing which content modifications increase a source's visibility inside generative search responses.

The post announcing the addition of Experience to Expertise, Authoritativeness and Trustworthiness, and clarifying that Trust is the central member of the E-E-A-T family.

Introduced RAG, the retrieve-then-generate pattern behind AI Overviews, ChatGPT search and Perplexity. Essential for understanding why source retrievability drives AI citation.

The most detailed technical analysis of the leaked Google Content Warehouse API documentation, covering NavBoost, click signals, site authority, twiddlers and document scoring modules.

The paper that introduced Google, describing PageRank, anchor text as a ranking signal, the crawling and indexing architecture and the repository design. The single most important document in SEO's intellectual history.

The post that made the leaked Google Content Warehouse API documentation public, and contrasted the attribute names found in it with Google's prior public statements.

Google's April 2012 announcement of the webspam algorithm update later named Penguin, which targeted link schemes and keyword stuffing and reshaped the link building industry.

Introduced the Transformer architecture, which underpins BERT, MUM, T5 and the large language models that now generate AI search answers.

Showed that learned dense embeddings can outperform BM25 for first-stage retrieval. The technical basis for vector search and for retrieval inside AI answer engines.

Google's announcement of the page experience signal and the introduction of Core Web Vitals as measurable components of it.

Introduced rel=sponsored and rel=ugc and reclassified nofollow as a hint rather than a directive, changing how paid and user-generated links are marked up.

Google's announcement of AI Mode, a conversational, query-fan-out search experience, alongside the wider rollout of AI Overviews.

The text-to-text framework Google cites as the architecture underlying MUM. Relevant to understanding multitask, multimodal and multilingual query handling in modern search.

Google's announcement of the general rollout of AI Overviews in the United States along with multi-step reasoning and planning features in Search.

Google's overview of generative AI in Search, describing the shift from a ranked list of links toward synthesised answers with supporting sources.

Google's confirmation that HTTPS is a lightweight ranking signal, which triggered industry-wide migration to TLS.

Google's introduction of the Multitask Unified Model, built on the T5 text-to-text architecture, trained across 75 languages and able to understand and generate language across modalities.

Demonstrated large gains from applying BERT as a re-ranker over retrieved passages on MS MARCO. Foundational to passage ranking as deployed in modern search systems.

The April 2015 announcement of mobile-friendliness as a ranking signal on mobile search, the update the trade press called Mobilegeddon.

The archive of Bill Slawski's patent analyses, which for two decades was the primary public source translating Google and Bing patents into SEO implications. Slawski died in 2022; the archive remains a reference body of work.

Google's announcement of BERT in Search, described at the time as the biggest leap forward in five years, affecting roughly one in ten English queries.

OpenAI's paper on training a model to browse the web and cite sources when answering. An early blueprint for the browsing and citation behaviour of current AI answer engines.

Moz's long-running weekly video series, begun by Rand Fishkin, explaining SEO concepts at a whiteboard. Collectively one of the most-cited free educational bodies of work in the industry.