17 replies to this topic

[1dmf]

I was reading a link from Randy regarding SE Stats and came across this http://searchenginel...0605-131904.php regarding new webmaster guidelines.

One of the rules says 'avoid hidden text and links' , now I'm a little confused with this. as I understood it a spider / bot / crawler is not capable of reading a web page, grabbing the CSS attached and then matching corresponding class->tag->CSS to see if it has visibility 'hidden' or display 'none'.

Or can they?

Also , there are a 101 reasons you might do this, drop down/pop out menus, tree structures, bespoke tooltip popups, etc... , how does the bot know if it's 'Bad SEO' or part of the site design.

And how does it affect your ranking if the bot decides it's a breach of the guidelines rather than site functionality.

This seem like an imposible and unfair requisite to try and meet.

Any thoughts peeps?

[BlueSky]

I was reading a link from Randy regarding SE Stats and came across this http://searchenginel...0605-131904.php regarding new webmaster guidelines.

One of the rules says 'avoid hidden text and links' , now I'm a little confused with this. as I understood it a spider / bot / crawler is not capable of reading a web page, grabbing the CSS attached and then matching corresponding class->tag->CSS to see if it has visibility 'hidden' or display 'none'.

... and setting text color to background color, etc. Of course bots can look at CSS... how else would this be detected? You have to assume that the bots can see all text and link trickery. This is only a concern for those who are up to no good.

[1dmf]

This is only a concern for those who are up to no good.

and how do you qualify this, how can it tell if something is hidden in context with site design and functionality or if it is "Cloaking" tactics.

Of course it's not only a concern for those up-to-no good, I supose you beleive spam filters never have "False Positives" then?

[BlueSky]

and how do you qualify this, how can it tell if something is hidden in context with site design and functionality or if it is "Cloaking" tactics.

Look at the page as rendered by the major browsers. If you can't see the text, then there is a problem.

Of course it's not only a concern for those up-to-no good, I supose you beleive spam filters never have "False Positives" then?

This is not the same as spam filters. You have control over what you give to the bots. Not to say that bots may interpret a page incorrectly, but considering the importance indexing quality, I have some measure of faith that they do it correctly. In other words, it is more likely you are concerned about something that does not deserve concern.

[qwerty]

I know a number of people believe that bots can read CSS and determine if you're using styles to hide anything, but I've never seen proof of it. Obviously, the search engines don't want you to hide text from people, and using CSS to temporarily hide text isn't the same thing, as there's no intention on your part to deceive anyone.

Does that mean you're perfectly safe using CSS to actually hide text? Not if someone reports you (supposedly), not if they really can read CSS (although I doubt it), and not if they eventually work out a way to detect it (which I think is quite likely).

But I think some well-intentioned practices are potentially dangerous. For example, I'm currently looking over a site that does its <h1> elements like this: the actual text of the heading is set by the CSS to be a few thousand pixels outside the browser window, and a background image of text is positioned where the heading should be. The intention here is clear: the designer wanted a specific look, so they used the image. Search engines can't read the text on the image, so they put the text there, but the text had to be kept invisible or it would sit on top of the background image.

But what happens to users with CSS on but images off? They don't see the background image or the text. If a search engine can figure out what's going on there, what conclusion is it going to come to?

[Mike Bradbury]

It is my experience that search engines look for red flags in CSS. Matching CSS to tags, I think, would be take up a LOT of bandwidth, and slow things down for the engine. So they don't do this.

However, I have seen search bots hit my CSS pages, and, based upon my research, they're looking for red flags. For example:

Pages I style with optimized text in tags that make use of styles such as:

left:-9999999px;
height:1px; overflow:hidden;

Have performed worse than pages with "height:25px; overflow hidden;"

The first two examples would pretty obviously be something hidden. I think this is what Google looks for. Based upon a
post by Matt Cutts: www.mattcutts.com/blog/seo-mistakes-unwise-comments/ as well, any inline "display:none" styles may also be suspect in the eyes of Google.

Furthermore, I don't think you'll be penalized outright for using methods such as these, because they still may be employed for legitimate purposes. I just don't think the content within tags styled using these "red flags" are weighted as heavily in the algorithm.

And I agree with Google for doing things this way. Anything a site visitor doesn't immediately see shouldn't be weighted as heavily as something that is always in plain site.

[1dmf]

Anything a site visitor doesn't immediately see shouldn't be weighted as heavily as something that is always in plain site.

intersting thought, but an example as i said is a link in a pop out menu, is it possible the page isn't indexed because on initial view of the page the link isn't visible, as it requires a hover effect to make it visible.

I'm not tying to hide anything or deceive no-one by using this CSS hover method, infact, it's the standards vs using JS to acheive the same effect, so , that said, does it look for a matching class that has a classname:hover {display:block}; to determine genuine hidden text on inital page view?

I just find this 'google guideline' to be very ambigous and nigh on impossible to evalute, I understand the priciples of SEO cloaking and why it's concidered bad, but link/text cloaking does not automatically mean bad SEO.

the two are not the same!

[Mike Bradbury]

In my experience, Google will index links that are found in drop-down menus. They also will not penalize you for this.


IF you're really that worried about it, just add the links in a sitemap. Google will find them there. Don't sacrifice the functionality of your site because google doesn't know the difference between spam and not spam.

[Randy]

I just find this 'google guideline' to be very ambigous and nigh on impossible to evalute

As it's meant to be. It gives them a lot more leeway in how they apply things inside the 'Plex, which is their right.

The main message Google has always given is Do Not Try To Fool Us. As long as you've got that one covered you're okay. So you can basically forget about worrying about anything else.

[BlueSky]

It is my experience that search engines look for red flags in CSS. Matching CSS to tags, I think, would be take up a LOT of bandwidth, and slow things down for the engine. So they don't do this.

What bandwidth? One, or perhaps a few CSS stylesheets for each website? That is not alot of bandwidth. Put yourself in Google's position -- you have the opportunity to fetch the CSS, you know how to interpret it, your #1 product is search quality, you want to stay at the top of your game, you don't want to be fooled, so what do you do?... you use the CSS. That's what I would do if I was responsible for search engine quality. We do not know to what extent this is implemented at the moment but I would bet my life that if not now, it is a sure thing eventually.

[Mike Bradbury]

What bandwidth? One, or perhaps a few CSS stylesheets for each website? That is not alot of bandwidth. Put yourself in Google's position -- you have the opportunity to fetch the CSS, you know how to interpret it, your #1 product is search quality, you want to stay at the top of your game, you don't want to be fooled, so what do you do?... you use the CSS. That's what I would do if I was responsible for search engine quality. We do not know to what extent this is implemented at the moment but I would bet my life that if not now, it is a sure thing eventually.

It just seems like a lot of stuff to take into account, and even if they are donwloading your stylesheets and matching them to to your HTML - CSS doesn't imply intention.

So, I'll stand by what I originally said - bandwidth aside - Google isn't going to penalize you for your 'hidden content' - and I know for a fact that they'll still index it.

[BlueSky]

It just seems like a lot of stuff to take into account, and even if they are donwloading your stylesheets and matching them to to your HTML - CSS doesn't imply intention.

So, I'll stand by what I originally said - bandwidth aside - Google isn't going to penalize you for your 'hidden content' - and I know for a fact that they'll still index it.

Not correct. Your site can be penalized for "hidden content":

"If your site is perceived to contain hidden text and links that are deceptive in intent, your site may be removed from the Google index, and will not appear in search results pages."

http://www.google.co...py?answer=66353

(if the link gets removed, enter "Google Help Hidden text and links" into Google.com)

[Randy]

I think you're talking about two different things folks.

Hidden content, when it is not deceptive and is simply a design element, will not get you penalized. I believe this is what Mike is referring to.

Hidden content that is clearly deceptive --white text on a white background, content positioned well off the screen and content that is hidden and cannot be made visible by user action-- may attract more scrutiny and penalty. And I believe this is what BlueSky is referring to.

Obviously, they're two totally different things. And for the record, it wouldn't be all that hard for Google to figure out if a hidden layer was coded in such a way that it could never be made visible. Childs play really, compared to everything else they do with the data they collect.

[Michael Martinez]

I know a number of people believe that bots can read CSS and determine if you're using styles to hide anything, but I've never seen proof of it.

http://www.mattcutts...nwise-comments/

Not the most definitive proof out there, but I don't feel up to doing a usual thorough search. Sorry. My sinuses are just killing me this week.

Um, in my opinion, if you do something sneaky, it won't matter if the search engine can algorithmically detect it. If any search engine's Web spam team (and all the majors have them, as I understand it) is asked to look at your site, you're doomed.

Best not to invite closer scrutiny by doing something deliberately wrong.

[qwerty]

Of course, we don't know how Matt found the site to use as his example. Did Googlebot find it? Did someone report it as spam?

In any case, I absolutely agree with the advice in your last sentence.