Is it Cloaking if Content is Identical

Changing meta tags for spiders, but not content

[r1000009]

An SEO company has told me this practice is accepted by Google, because it does not change the page or user experience, only other elements. I would be keen to hear other peoples views on this.

When a spider visits the site and follows a link, an alternative, static, version of the page is returned. On this page the meta tags, title and URL are all modified to be more SE friendly.

They also tell me a number of big companies are doing this already and have not been penalised. Does this sound like it would genuinely be accepted by Google, or is it only a matter of time until they clamp down on it?

[volatilegx]

Welcome to WebmasterWorld, r1000009 :)

Does this sound like it would genuinely be accepted by Google, or is it only a matter of time until they clamp down on it?

That is a question that is subject to much debate. I believe you could get away with such for a long time without being penalized.

[r1000009]

many thanks.

[DamonHD]

Hi,

I avoid sending meta tags to non-spider visitors since it simply wastes bandwidth and time (users don't read meta tags; spiders sometimes do).

No attempt is being made to deceive anyone, so no one should mind.

I also use slightly less CPU-expensive stuff on pages when I think the visitor is a spider since it can save me 0.25s of CPU time building the page and the difference is marginal (actually, I also give the spider a very slightly more focussed set of links off the page to related items in the site as a result).

Again, no attempt to deceive nor material change, so no one should mind. And I really don't mind if normal humans see the "spider" version of a page; it's totally usable and presentable.

Rgds

Damon

[Brett_Tabke]

I completely agree with you DamonHD. I too save robot content for the robots (meta tags) and don't serve that to humans and save mega bandwidth. The question is one of intent. As long as the page reflects the content, SE's are not going to mind at all.

[volatilegx]

it can save me 0.25s of CPU time building the page

... but how much CPU time does it take to decide if the visitor is a spider?

[DamonHD]

Hi,

At a guess, about 100th of that or less! Pretty well too small to measure.

Rgds

Damon

[martingale]

I can't believe meta tags consume enough bandwidth that anybody cares. You better not serve any flash on your site if you're going to say anything like that.

[piskie]

Brett, if BW is an issue, I think you would be better off getting rid of font tags and table tags before you start squeezing that last ounce out of meta tags.

Maybe you could use CSS which is unquestionably OK with S/Es whereas your white cloaking may not always be if indeed it is now.

<DUCK>

[DamonHD]

MG,

No, I don't like Flash; I'm not a fan of Flash at all.

My site guidelines aim for most pages to be under 32kBytes, *including* all embedded content such as inline images.

Thus, saving a few percent of metatags helps with bandwidth (more so in the past when it was much more expensive for me).

But much more importantly it significantly reduces the time before the first viewable and renderable content reaches the viewer's browser: one or two TCP segments at least which may *double* the perceived performance on a slow or congested link. So the user sees something sooner AND the browser can start fetching embedded images sooner.

So the motivating factor is as much UI experience and interactivity as bandwidth.

I learnt all this the hard way many many years ago when I got into Mirskey's "Worst of the Web" and my 14400bps analogue leased line was maxed out for a whole week! B^>

Rgds

Damon

PS. I had a quick sanity check regarding earlier posts in the thread, and almost all the search engines I care about, esp G, do have no "referrer" so I will stick with this very quick check for the moment.