Free field guide
The 6 Papers behind modern SEO
The Google patents and research papers that actually drive rankings — each one explained in plain English, with the concrete SEO takeaway. No gurus, no opinions.
Read top to bottom — they build on each other, from how authority flows between pages to how Google understands meaning across text, images, and video.
- 01Patent US 6,285,999Stanford / Google · 1998
PageRank
PR(A) = (1−d)/N + d · Σ PR(i)/L(i)
The mathematics of how authority flows between pages. The basis of our internal linking and link strategy.
Internal linkingAuthorityArchitecture - 02“Things, not strings”Google Search · 2012
Knowledge Graph
entity → relation → entity
The shift from keywords to entities. Shapes our schema.org markup and entity-based topic clusters.
EntitiesSchema.orgClusters - 03arXiv 1301.3781Google Brain · 2013
Word2Vec
cos(“seo”, “ranking”) ≈ 0.21
Words as vectors in a continuous space. The foundation of our semantic proximity and co-occurrence analysis.
EmbeddingsSimilarityLexical field - 04arXiv 1706.03762Google Brain · 2017
Transformer
Attention(Q,K,V) = softmax(QKᵀ/√d)·V
Multi-head attention — the architecture behind BERT, MUM and modern Google Search.
AttentionArchitectureLLM - 05arXiv 1810.04805Google AI · 2018
BERT
context: left ↔ right
Bidirectional context understanding. Informs our content briefs and on-page semantic structure.
NLPEmbeddingsBriefs - 06Google I/OGoogle Research · 2021
MUM
text + image + video → answer
Multimodal, multilingual understanding. Guides our work on rich SERPs and hreflang.
MultimodalHreflangRich SERP
Read the primary sources:
This is one issue. The briefing ships every week.
If this guide was useful, the weekly briefing turns the latest patents and papers into patterns you can deploy. Or, if you'd rather we did it for you, book a free audit.