It's serving a different page to the search engines than you serve to the user (surfer). There are lots of ways to do it.
cloaking is black hat techniques, means your real content only will show for search engines not your visitors. http://search.msn.com/results.aspx?q=cloaking http://www.webmasterworld.com/forum24/835.htm
Cloaking is absolutely legal. Whether it is against a search engine's terms of service is another issue entirely. But it's not like you'd get arrested for it. You may, however, get delisted by Google. From what I know: Yahoo: "Cloaking is OK so long as it's not used deceptively." MSN: "Ditto." Google: "Cloaking is evil, but it's OK for the New York Times for some reason " But it's definately "legal." My gut feeling is that cloaking, so long as you're not substantially changing the content, is not something you're going to get in trouble for, even with Google. What the New York Times does, in my opinion, doesn't quality for that definition. Doing stuff like changing a few graphical headers to text, and in general leaving the rest of the content the same, is probably safe -- and in my opinion, this helps the spiders. It's all about intent. I'd avoid redirects and fundamentally different content, though. And I'd use both agent and IP detection to detect a spider. IMO, there are legitimate uses for cloaking. The implementation of IP delivery and cloaking is the same. Again, it's all about intent. Strictly speaking, it's still against Google's TOS, but apparently Google has been looking the other way for a few larger content providers ... But if you get banned, it's still your problem And I'd still avoid it if there are other alternatives (there usually are). -J.