01
Trace every number
Every number traces to a primary source or a stated method.
01
The AI-visibility
research desk.
Shareof.ai research desk · Vol. 01 · Jul—Aug 2026The AI-Visibility
The AI-Visibility
Content Engine.
Original frameworks, technical field notes, and reproducible methods for understanding how AI systems retrieve, cite, and recommend brands.
02
Connect every framework
Every framework links to the evidence and the next practical step.
03
Separate the evidence
Every study separates observations, inference, and limitations.
01Flagship field paper · August 26, 2026
How AI Models Actually Choose Which Brands to Cite
A technical teardown of the retrieval pipeline, and why the metric you should be tracking sits one step upstream of the one everyone talks about.
13 min readPillar research
Read the paper ↗Research archive
The complete field library
01Pillar research
How AI Models Actually Choose Which Brands to Cite
A technical teardown of the retrieval pipeline, and why the metric you should be tracking sits one step upstream of the one everyone talks about.
19Technical playbooks
Reverse-Engineering Co-Citation: Find the Pages That Decide Your Category
If you want on the lists an AI hands people, first find the handful of pages it reads before it answers. Here is how to rank them.
02Pillar research
What the Princeton GEO Paper Really Proved (and Where Practitioners Get It Wrong)
Everyone quotes the "+40%" number. Almost nobody read the method that produced it, so the advice built on top of it keeps misfiring.
03Pillar research
How ChatGPT Search Actually Works: The Pipeline Behind Every Answer
A stage-by-stage walkthrough of what happens between your question and the cited answer, and where your content wins or loses at each step.
10Contrarian theses
'AI SEO' Is Optimizing the Wrong Layer
Most GEO advice tunes the content that gets cited. The harder question is whether your page was ever in the room when the model chose.
14Core frameworks
Share of Voice vs Share of Model vs Share of Retrieval
Three metrics get thrown around as if they were the same number. They measure different stages of the same process, and confusing them is why so many AI-visibility dashboards feel busy but tell you nothing.
15Core frameworks
The AI Visibility Funnel: A Diagnostic for Where You're Losing Citations
Four stages sit between your page and a named mention in an AI answer. Find the one that is failing before you spend a week fixing the wrong layer.
16Core frameworks
Map Your Citation Surface: Every Page Where AI Can Find You
Most of the time an AI names your brand, it read that name on someone else's page. Here is how to find and fix every one of those pages.
12Contrarian theses
Your "AI Rank" Is a Vanity Metric
A single number pulled from one run of a stochastic system is a screenshot, not a measurement. Here is what to track instead.
11Contrarian theses
Stop Tracking Keywords. Start Tracking Prompts.
The keyword was a unit built for a search box that no longer exists. Here is what replaces it.
13Contrarian theses
Schema Markup Won't Save You: What Structured Data Does and Doesn't Do for LLMs
Schema.org markup earns its place for entity clarity and search eligibility. It is not the citation lever the AI-visibility pitch decks say it is.
24Buyer guides
What to Do When a Competitor Owns the AI Answer for Your Category
A staged counterattack for the moment you keep watching ChatGPT and Perplexity recommend a rival by name while you never come up.
22Buyer guides
How to Choose an AI Visibility Tracking Tool (An Honest Buyer's Guide)
A real evaluation rubric for buyers who want to know how ChatGPT, Claude, Gemini, and Perplexity see their brand, plus a scorecard you can apply to any vendor.
21Technical playbooks
A/B Testing Content for AI Citation: A Methodology That Survives the Noise
Ask an AI the same question twice and you may get two different sets of sources. Any before-and-after test that ignores this is measuring noise and calling it a result.
18Technical playbooks
Build a Brand-Mention Monitor Across Four LLMs in an Afternoon
A working script that runs your prompt set against ChatGPT, Claude, Gemini, and Perplexity on a schedule, then logs where your brand shows up and who gets cited.
17Technical playbooks
Measure Your Own Share of Retrieval With the Model APIs
Most people test whether AI cites their site by typing their brand into ChatGPT once. Here is how to measure it properly, with code you can run today.
23Buyer guides
The 15-Minute AI Visibility Audit You Can Run Right Now
A timed, do-it-yourself pass that gives you a real read on whether ChatGPT, Claude, Gemini, and Perplexity name your brand, and points at the exact stage where you are losing.
20Technical playbooks
Reading Your AI Crawler Logs: GPTBot, ClaudeBot, PerplexityBot and Friends
Your access log already knows which AI engines can see you. Here is how to read it, and why the live-fetch agents matter more than the training crawlers.
See which part of the visibility funnel is losing your brand.
Run a free scan across real buyer questions, then use the field library to decide what to fix first.
Run the free scan ↗