Cloaking Question

[TenTonJim]

Hello everyone,
First post here at WebProWorld. I am also over at the sitepoint forums http://sitepoint.com.

I know ethically that cloaking is considered bad seo practice, but how does a spider know that is it getting cloaked pages?

Say I am using ASP, and am determining the presence of MSIE, FireFox, Gecko, Opera etc etc in the HTTP_USER_AGENT string when a visitor arrives at the site, and feeding browsers visual content and spiders gratuitous text... how does the spider know this? It seems tht only dmoz or perhaps yahoo would know it was being cloaked.

I don't practice this but am curious as to how a spider determines cloaking.

Thanks!

[fathom]

Originally Posted by TenTonJim

Hello everyone,
First post here at WebProWorld. I am also over at the sitepoint forums http://sitepoint.com.

I know ethically that cloaking is considered bad seo practice, but how does a spider know that is it getting cloaked pages?

Say I am using ASP, and am determining the presence of MSIE, FireFox, Gecko, Opera etc etc in the HTTP_USER_AGENT string when a visitor arrives at the site, and feeding browsers visual content and spiders gratuitous text... how does the spider know this? It seems tht only dmoz or perhaps yahoo would know it was being cloaked.

I don't practice this but am curious as to how a spider determines cloaking.

Thanks!

"Ethically" would be a bad word choice - as it implies deception.

Google deploys cloaking technology to redirect users to their regional version of Google, as well as limiting Adwords from displaying in results for say a user on Google.com but located in UK where UK was not selected as an ad placement. In both instances Google isn't being deceptive.

Notwithstanding, whether used for the right (or wrong) reason the user agent can't tell (unless specified and even then Googlebot would unlikely know this as a deception tactic) - thus if used for the wrong reason is the only concern and in this instance a spam report would be needed to get Google to take a closer look.

I would add though - this is of my opinion - I am not professionally versed in cloaking technology just using common sense.

[TenTonJim]

Thank you for your comments.

More questions though...

How would a user create a spam report if they only see the graphical representation of the site?

Would they deduce that it was cloaked by comparing the search engine results to the actual site?

A bit further, it could probably be done so that they mimick one another, yet the cloaked one presented in such a way as to maximize pure html tags with little representational code... somewhat like a plain text version of a website.

Thoughts on this?

Thanks!

[DeYo]

There is no way for spiders to "know" which page is cloaked and which one is not! You have to be careful though because your competitors might notice your tactics and report it back to Google.

[TenTonJim]

I am being totally serious.

Please, I am not being argumentive just looking for a definitive answer. I have read a lot about being banned for cloaking and was just wondering how that happens. I understand if you make it an obvious spam page... but what if you don't make it obvious?

How would they know?

If it appears in the SE results as the text of the graphical site, yet the plain text version is stripped of almost all html, how could anyone figure that out?

Thanks for your comments!

[TenTonJim]

Someone just answered this for me over at sitepoint.

The Google cache.

It is a snapshot of what Google "sees". So it would be obvious that it was a plain text version of what was shown to the public.

Thanks for all of your comments!

[fathom]

additionally you can on mouseclick not release on link, move off link area and release. The cloaked page can be revealed.

[scisoft]

detecting HTTP_USER_AGENT is not the way to do this. If Opera can pretend being IE, spiders can. The easiest way to detect cloaking is two spiders coming form two diferent IPs at the same time. One of them identifies itself as IE or other common browser. If they do not get the same content, you are in a trouble. It is known that spiders do this.
You need to know all spiders IPs. I found a spider list there http://www.spiderhunter.com/ but I bet it is not complete.