Not so-much related to cloaking, Disney.com had to undergo a massive strategy for SEO when the news came down that they’d be rolling out an all-Flash website. Fortunately, the company hired my friend Mei Huang (http://twitter.com/meihuang) who helped insure the right amount of scripting went into the architecture to insure that the new wonderful UI didn’t hose years of optimization work. As I recall, she was in constant communication with Google, Yahoo! and Microsoft, to insure that Disney followed the rules. Cloaking wasn’t involved, but a whole heck of a lot of JavaScript was.

If a project is a big enough to warrant the need for cloaking by user-agent, it might be prudent to reach out directly to Google for guidance. Maybe even a quick ping to http://twitter.com/mattcutts might be in order. I’m sure he appreciate the fact that a business is asking before jumping to a conclusion that could get a site penalized.

Either way, the risk is worded by Google as “Serving up different results based on user agent may cause your site to be perceived as deceptive and removed from the Google index.”

The link:
http://www.google.com/support/webmasters/bin/answer.py?hl=en&answer=66355

They also identified an approach to handle these types of scenarios, “Ensure that you provide the same content in both elements (for instance, provide the same text in the JavaScript as in the noscript tag). Including substantially different content in the alternate element may cause Google to take action on the site.”

So as you mentioned, there are good and bad reasons to use the technique, but if you are still returning the same content (in one form or another), you have less of risk of penalty.

Thanks so much for participating by the way, it’s an honor to have you here. )

This is a good start but the cloaking discussion is a complex one. There are people who use these technologies to benefit both users and search engines.

For example, some entertainment companies show content in one country that is not available in other countries. They have to use IP detection to make a best guess about who should and should not have access to their content.

And there are other pro-cloaking arguments, some of which may be a bit dated because new technologies make it possible to optimize the user experience and the search experience separately without swapping pages.

I think any standard dealing with this complex issue has to focus on both intent and the end result. If the intent is to deceive the users with a goal of depriving them of a fair experience, then it’s not acceptable. But in some circumstances the intent is to abide by legal or contractual obligations or to ensure that users find what they are indeed looking for.