Articles
Research, experiments and analysis on AI visibility, generative search, SEO and how artificial intelligence systems understand the web.
Google Search, AI Overviews and Gemini Often Choose Almost Completely Different Sources for the Same Query
Google Search, AI Overviews and Gemini may all belong to the same company, but they frequently construct their answers from strikingly different parts of the web. A large comparati...
Read more →
Nearly 30% of Google AI Overview Sources Didn’t Rank on Page One—and 11% of Claims Lacked Citation Support
Google AI Overviews do not simply summarize the websites visible at the top of the same search results page. A large-scale measurement study found that 29.8% of domains cited by AI...
Read more →
GEO Rewriting Can Make a Page More Citable—and Remove It From Retrieval at the Same Time
A page can become easier for a language model to cite and simultaneously become less likely to reach that language model at all. That is the central warning from SAGEO Arena, a new...
Read more →
Google AI Mode Cut External Click-Through by 18.8 Points—and Pushed More Users Toward Rival Search Engines
A preregistered field experiment with 1,100 Google users has produced unusually strong causal evidence for a concern publishers have mostly had to measure through observational tra...
Read more →
Fashion Brands Now Have an AI Visibility Score Alongside Their Traditional Media Rankings
Fashion has spent years turning media attention into rankings, dashboards and monetary-style impact scores. Now the same measurement machinery is moving into generative AI. Launchm...
Read more →
AI Engines Changed Their Social Citation Mix 16 Times in Seven Months—and Announced None of It
Social media is becoming a larger evidence layer for AI answers, but the platforms that benefit can change in a matter of weeks. Goodie says it tracked 16 distinct shifts in social...
Read more →
Google Is Testing Bigger Citation Cards—but Moving Them to the Bottom of AI Mode
Google appears to be testing a larger, more descriptive source panel in AI Mode while simultaneously moving those citations farther away from the generated answer. The experimental...
Read more →
OpenAI’s New API Lets Anyone Build Long-Running AI Search Agents With Web Retrieval and Subagents
OpenAI is making the infrastructure behind its long-running agents available to developers. The new Agents API, now in public beta, packages the agent harness and cloud infrastruct...
Read more →
ChatGPT Turns Governed Company Data Into a Searchable Answer Engine With Inspectable Evidence
OpenAI is turning ChatGPT Work into a conversational analytics layer for company data. Its new Data agent can connect to approved warehouses, databases, documents and business-inte...
Read more →
Google Will Preview Its Next Local AI Search Updates on September 24
Google representatives are scheduled to preview upcoming AI Search changes affecting local discovery during a September 24 webinar, giving local marketers a date to watch for new i...
Read more →
The Same Anthropic Crawl-to-Referral Metric Produced Seven “Correct” Answers
Ask for Anthropic's crawl-to-referral ratio and you can get seven dramatically different answers: 70,900:1, 38,000:1, 23,951:1, 11,122:1, 10,300:1, 4,580:1 or 2,237:1. All seven ha...
Read more →