Generative Engine Optimization, grounded in evidence.
Original research, practical guides and documented comparisons. Clear methods, open datasets and limits stated beside the findings.
Latest publications
Compare AI visibility tools →- measurement · 3 Oct 2026Cornerstone
AI Visibility Tools Compared: 35 Platforms, Features and Prices
We read the pricing pages and documentation of 35 AI visibility trackers on 3 October 2026 and compared them on engines, method, metrics, actions, integrations, compliance and price. 12 in-depth reviews, no rankings, and the full dataset to download.
- measurement · 7 Jun 2026Cornerstone
State of GEO Q2 2026: the AI engine you optimize for matters most
We probed 100 brands across ChatGPT, Claude, Gemini, and Perplexity — 4,000 prompts. Citation rate varies far more by which engine you ask than by which brand you are: Perplexity cites 42%, ChatGPT 7%, Gemini hides its sources, and legacy news publishers get 0%.
- measurement · 7 Jun 2026
Who LLMs cite for GEO: 927 sources, 417 domains, zero of them us
Across 50 GEO questions, four LLMs returned 1,105 source links. After setting aside Gemini's opaque redirects, the 927 attributable citations spread across 417 domains — a long tail led by YouTube and SEO-tool blogs. Here's the map, with the grounding caveat up front.
- technical · 7 Jun 2026
JSON-LD Recipes for Articles, Datasets, and FAQs
Copy-paste JSON-LD blocks that pass schema.org validators, render correctly in Google Rich Results, and surface in AI search engines. Annotated with what each field actually does.
- case-studies · 31 May 2026Cornerstone
This site is our GEO lab: the stack, the data, the experiments
A dated account of our GEO lab: the historical June 2026 baseline, later citation coverage and the limits of crawler identity verification. Old measurements remain history, not current performance claims.
- foundations · 31 May 2026
GEO vs AEO vs LLMO vs SGE: An Honest Taxonomy
Four acronyms, mostly the same thing — and a few clear distinctions worth keeping. We pulled the original definitions, checked who uses which term, and argue which one should win.
Recently updated
- AI Visibility Tools Compared: 35 Platforms, Features and PricesUpdated 21 Oct 2026 · published 3 Oct 2026
- State of GEO Q2 2026: the AI engine you optimize for matters mostUpdated 21 Oct 2026 · published 7 Jun 2026
- Who LLMs cite for GEO: 927 sources, 417 domains, zero of them usUpdated 21 Oct 2026 · published 7 Jun 2026
Pillars
Foundations
Theory, mechanics, and history of AI search. How LLMs retrieve, rank, and cite.
Technical
Server-side, code-level, infra-level. llms.txt, robots, schema, structured data.
Measurement & Analytics
Tools, metrics, methodology. Reproducible benchmarks.
Case Studies
Brand-by-brand teardowns. Sector benchmarks.
Recommended starting points
- measurement
AI Visibility Tools Compared: 35 Platforms, Features and Prices
We read the pricing pages and documentation of 35 AI visibility trackers on 3 October 2026 and compared them on engines, method, metrics, actions, integrations, compliance and price. 12 in-depth reviews, no rankings, and the full dataset to download.
- measurement
State of GEO Q2 2026: the AI engine you optimize for matters most
We probed 100 brands across ChatGPT, Claude, Gemini, and Perplexity — 4,000 prompts. Citation rate varies far more by which engine you ask than by which brand you are: Perplexity cites 42%, ChatGPT 7%, Gemini hides its sources, and legacy news publishers get 0%.
- case-studies
This site is our GEO lab: the stack, the data, the experiments
A dated account of our GEO lab: the historical June 2026 baseline, later citation coverage and the limits of crawler identity verification. Old measurements remain history, not current performance claims.
- technical
llms.txt Adoption: Our 100-Domain Audit and Its Limits
Our May 2026 audit found llms.txt at 37 of 100 sampled apex domains. Preserve the historical evidence, understand what it cannot prove, and follow the updated proposal.
- foundations
GEO Methods: An Evidence-Led Taxonomy
Five groups of GEO methods, with clear boundaries between documented mechanisms, research findings and hypotheses. Practical steps without guaranteed citation gains.