AI Browser Guide: How Agents Read Your Site
What an ai browser can actually read and click on your site: forms, menus and logins, plus a 10-minute self-test you can run on your own pages.
How ChatGPT, Perplexity, Gemini and Google AI Mode retrieve, rank and cite sources.
41 articles, newest first
What an ai browser can actually read and click on your site: forms, menus and logins, plus a 10-minute self-test you can run on your own pages.
What happened to the search generative experience (SGE): the timeline from Labs test to AI Overviews and AI Mode, and which SGE-era advice still holds.
NLWeb explained without the hype: what Microsoft's open project documents, what it does not prove, and a plain test for whether your site should adopt it now.
CCBot feeds Common Crawl, a dataset behind many AI models. See what blocking it changes, what it does not, and how to verify real CCBot visits in your logs.
GPTBot, OAI-SearchBot and ChatGPT-User do three different jobs. See which one controls ChatGPT search visibility and the robots.txt rules for each goal.
ClaudeBot is Anthropic's training crawler, one of three Claude bots. See what each does, the robots.txt rules to control them, and how to check your logs.
Answer engine optimization tools sorted into six types by the job they do, what each cannot verify, and a free baseline test to run before you pay.
AI search optimization tactics ranked by effort and by their effect on retrieval versus citation, plus a stop-doing list checked against Google and OpenAI docs.
What IndexNow is, which search engines use it, how to create the key file and send the API call, plugin options, and what IndexNow cannot do for Google.
Claude web search from Anthropic's own docs: how to turn it on, how it decides to search and cites, which crawlers it uses, and how to get your pages cited.
Featured snippets explained: the formats, how Google picks them, the page structure that wins them, how they relate to AI Overviews, and how to opt out.
ChatGPT shopping explained from OpenAI's own pages: how products and merchants are picked, what feeds and reviews do, and what merchants can change right now.
Bing SEO from Bing's own documents: the six ranking parameters, where Bing differs from Google, the directives that change Copilot citations, and Bing Places.
Search engines without AI, checked against their own help pages: Google's Web filter and udm=14, DuckDuckGo No-AI, Mojeek, Kagi, and what it means for sites.
YouTube SEO: titles, descriptions, chapters, transcripts, thumbnails, VideoObject markup on your site, and why YouTube is the top source in AI Overviews.
ChatGPT deep research explained from OpenAI's docs: how it plans, browses and cites, limits by plan, using it for marketing research, and how to get cited.
What Google's people also ask box is, where its answers come from, how to mine it for a content plan, and why PAA is the best free preview of AI fan-out.
AI search engines compared from their own docs: which index each one searches, which crawler feeds it, how it cites, and what that means for your website.
A practical Bing Webmaster Tools guide: import from Search Console, verify, submit sitemaps and IndexNow, read the reports, and use the AI Performance data.
Semantic search explained: embeddings, vector similarity and hybrid retrieval, Google's path from Hummingbird to BERT and MUM, and what it means for content.
Plain-English definitions of 56 AI search terms, from AEO and query fan-out to grounding, RAG, sameAs, and share of model, with links to full guides.
A 40-point GEO checklist for AI search visibility: access and rendering, entity clarity, quotable content, off-site corroboration, and measurement.
Perplexity retrieves before it writes. Understanding the four stages between a question and a citation explains why good pages still get left out.
AI citation tracking measures how often AI engines name and quote your brand. What it measures, why it isn't rank tracking, and how it's really done.
AI engines cite specific pages, not whole sites. Three signals show you which of yours get quoted — referral data, crawler logs and sampled prompt runs.
AI agents are moving from answering questions to completing tasks: comparing, booking, buying. Here's what that changes about how your site needs to work.
Google AI Mode explained: how query fan-out works, why it differs from AI Overviews, and the practical steps to become a source it pulls into answers.
AI engines retrieve far more pages than they cite. Only 15% of retrieved pages reach the answer. Here's what happens in between and how to survive the cut.
AI engines send traffic but hide the attribution. How to build AI referral reporting in GA4, what Search Console shows, and what stays invisible.
AI referral traffic is concentrated in ChatGPT, with Gemini now past Perplexity. The March 2026 shares, the crawl-to-referral gap, and what to do about it.
How AI engines pick sources in 2026: the two-stage retrieval and selection model, what the research shows correlates with citation, and what does not transfer.
How much do AI Overviews cut clicks? The studies show a 34 to 61% range, zero-click at 68%, and about half the decline predating AI Overviews entirely.
llms.txt explained: what the proposed file is, who proposed it, whether any AI engine actually reads it yet, how to build one, and when to skip it.
How to appear in Google AI Overviews: what triggers them, how they relate to ranking, the structure that gets extracted, and how to measure appearances.
An engine-by-engine method for tracking brand mentions in Perplexity, Gemini, Google AI Overviews and AI Mode, with the fields to log and how they differ.
AI citation tracking compared: manual logging, scripted API runs, and paid platforms, what the tool landscape looks like, and how to evaluate any vendor.
Perplexity cites sources on nearly every answer. How its own index and crawlers work, what gets quoted, and the steps to become one of the cited pages.
How to rank in ChatGPT: how its search actually retrieves and selects sources, why Bing matters, and the steps that make your brand one it quotes.
A repeatable method for measuring ChatGPT visibility: designing a prompt set, controlling the run, what to log, counting rules, and turning it into a cadence.
How to get cited by ChatGPT and Perplexity: how each engine picks sources, what the citation research shows, and a 10-step playbook with a checklist.
What GEO means: generative engine optimization explained, how it differs from SEO, who it is for, what the work involves, and how to tell if it is working.
Every engagement starts with a free visibility audit, so we agree on the baseline before anyone writes code.