PLATFORMS
Platform Playbooks
Per-platform citation mechanics, crawler access, optimization checklists and measurement.
32 pages
How ChatGPT Decides What to Cite
Getting cited in ChatGPT requires content to be inside its retrievable set and, for a given question, more worth citing than the alternatives.
Updated 2026-08-06ChatGPT Crawler Access
ChatGPT's retrieval crawlers are OAI-SearchBot, ChatGPT-User and its training crawler is GPTBot; citation requires allowing the retrieval ones, while training access is a separate brand decision.
Updated 2026-08-06ChatGPT Optimization Checklist
Optimising for ChatGPT runs in four layers: fix crawlability first, then content and evidence, and finally third-party sources.
Updated 2026-08-06Tracking Brand Visibility on ChatGPT
Tracking ChatGPT means measuring the distribution of a fixed question set across repeated samples — not the position of any single result.
Updated 2026-08-06How Gemini Decides What to Cite
Getting cited in Gemini requires content to be inside its retrievable set and, for a given question, more worth citing than the alternatives.
Updated 2026-08-06Gemini Crawler Access
Gemini's retrieval crawlers are Googlebot and its training crawler is Google-Extended; citation requires allowing the retrieval ones, while training access is a separate brand decision.
Updated 2026-08-06Gemini Optimization Checklist
Optimising for Gemini runs in four layers: fix crawlability first, then content and evidence, and finally third-party sources.
Updated 2026-08-06Tracking Brand Visibility on Gemini
Tracking Gemini means measuring the distribution of a fixed question set across repeated samples — not the position of any single result.
Updated 2026-08-06How Google AI Overviews and AI Mode Decides What to Cite
Getting cited in Google AI Overviews and AI Mode requires content to be inside its retrievable set and, for a given question, more worth citing than the alternatives.
Updated 2026-08-06Google AI Overviews and AI Mode Crawler Access
Google AI Overviews and AI Mode's retrieval crawlers are Googlebot and its training crawler is Google-Extended; citation requires allowing the retrieval ones, while training access is a separate brand decision.
Updated 2026-08-06Google AI Overviews and AI Mode Optimization Checklist
Optimising for Google AI Overviews and AI Mode runs in four layers: fix crawlability first, then content and evidence, and finally third-party sources.
Updated 2026-08-06Tracking Brand Visibility on Google AI Overviews and AI Mode
Tracking Google AI Overviews and AI Mode means measuring the distribution of a fixed question set across repeated samples — not the position of any single result.
Updated 2026-08-06How Perplexity Decides What to Cite
Getting cited in Perplexity requires content to be inside its retrievable set and, for a given question, more worth citing than the alternatives.
Updated 2026-08-06Perplexity Crawler Access
Perplexity's retrieval crawlers are PerplexityBot, Perplexity-User and its training crawler is not separately identified; citation requires allowing the retrieval ones, while training access is a separate brand decision.
Updated 2026-08-06Perplexity Optimization Checklist
Optimising for Perplexity runs in four layers: fix crawlability first, then content and evidence, and finally third-party sources.
Updated 2026-08-06Tracking Brand Visibility on Perplexity
Tracking Perplexity means measuring the distribution of a fixed question set across repeated samples — not the position of any single result.
Updated 2026-08-06How Microsoft Copilot Decides What to Cite
Getting cited in Microsoft Copilot requires content to be inside its retrievable set and, for a given question, more worth citing than the alternatives.
Updated 2026-08-06Microsoft Copilot Crawler Access
Microsoft Copilot's retrieval crawlers are Bingbot and its training crawler is not separately identified; citation requires allowing the retrieval ones, while training access is a separate brand decision.
Updated 2026-08-06Microsoft Copilot Optimization Checklist
Optimising for Microsoft Copilot runs in four layers: fix crawlability first, then content and evidence, and finally third-party sources.
Updated 2026-08-06Tracking Brand Visibility on Microsoft Copilot
Tracking Microsoft Copilot means measuring the distribution of a fixed question set across repeated samples — not the position of any single result.
Updated 2026-08-06How Claude Decides What to Cite
Getting cited in Claude requires content to be inside its retrievable set and, for a given question, more worth citing than the alternatives.
Updated 2026-08-06Claude Crawler Access
Claude's retrieval crawlers are Claude-SearchBot, Claude-User and its training crawler is ClaudeBot; citation requires allowing the retrieval ones, while training access is a separate brand decision.
Updated 2026-08-06Claude Optimization Checklist
Optimising for Claude runs in four layers: fix crawlability first, then content and evidence, and finally third-party sources.
Updated 2026-08-06Tracking Brand Visibility on Claude
Tracking Claude means measuring the distribution of a fixed question set across repeated samples — not the position of any single result.
Updated 2026-08-06How Doubao Decides What to Cite
Getting cited in Doubao requires content to be inside its retrievable set and, for a given question, more worth citing than the alternatives.
Updated 2026-08-06Doubao Crawler Access
Doubao's retrieval crawlers are Bytespider and its training crawler is Bytespider; citation requires allowing the retrieval ones, while training access is a separate brand decision.
Updated 2026-08-06Doubao Optimization Checklist
Optimising for Doubao runs in four layers: fix crawlability first, then content and evidence, and finally third-party sources.
Updated 2026-08-06Tracking Brand Visibility on Doubao
Tracking Doubao means measuring the distribution of a fixed question set across repeated samples — not the position of any single result.
Updated 2026-08-06How DeepSeek Decides What to Cite
Getting cited in DeepSeek requires content to be inside its retrievable set and, for a given question, more worth citing than the alternatives.
Updated 2026-08-06DeepSeek Crawler Access
DeepSeek's retrieval crawlers are no published identifier and its training crawler is not separately identified; citation requires allowing the retrieval ones, while training access is a separate brand decision.
Updated 2026-08-06DeepSeek Optimization Checklist
Optimising for DeepSeek runs in four layers: fix crawlability first, then content and evidence, and finally third-party sources.
Updated 2026-08-06Tracking Brand Visibility on DeepSeek
Tracking DeepSeek means measuring the distribution of a fixed question set across repeated samples — not the position of any single result.
Updated 2026-08-06