Cloaking
Serving different content to search engines and to human visitors, usually by detecting the user agent or IP address of the requester.
Analogy Mental model
Laying out fresh beef and sterile gloves when the inspector calls, then serving the spoiled meat to the paying customers the moment the door closes.
In practice What you actually do
This is treated as the most serious category of violation, so most working risk is accidental: a broken prerender cache serving an old version to crawlers, a misconfigured paywall, or a personalisation layer that returns nothing to a bot. Test what a crawler actually receives.
Where people get burned Pitfalls
If content sits behind a subscription, declare it honestly with the appropriate structured data flagging that the work is not freely accessible. That is the sanctioned mechanism; simply letting crawlers past the wall while blocking readers is not.
This entry was written in English from the Chinese original, not machine-translated. Read the Chinese version →