RSS Feed
Twitter
September 8th, 2005 
QuadsZilla Some site owners (blakhat and whitehat) are stingy with outbound links. They think that because the 1998 Google Algorythm paper implied that each outbound link bleeds pagerank, that this still applies today.
It does not.
In fact, our data and testing suggests that linking to 2-4 external subject authorities actually increases a pages credibility with the search engines.
How do you find subject authorities? For white hats, you should know what sites are relevant to your niche and link to them liberally.
For black hat seo, your site building script should include 1-4 links per page to subject authorities. Obviously, you want to automate this process.
These authorities could be:
1. Random search result from the top 50 Google results for the keyword phrases.
2. Links to the wikipedia
3. dmoz results
4. Old media news outfits (New York Times, Forbes.com . . . etc.)
Obviously, since you don’t want people clicking on these links instead of your PPC or affiliate program, you have several options. These links should be linearly (on page code) very close to your body text. However, you can uses the CSS to put the division with these links into cold zones on the page – or make them disappear entirely.
The white hat way to go is to use the CSS to put these links in a cold portion of the page. This article from Google includes this diagram:
The white zones are read least.
Here are two tips from Captain Obvious:
1. Don’t make these nofollow links – that defeats the purpose of linking to authorities.
2. Make the links target=blank so that if one of your surfers clicks it, there is still a chance that they will click on one of your ads too.
For the darker hats, you can use the following in your CSS to make a division invisible:
#important {
visibility: hidden;
}
or
#body{
display:none;
}
or
#main{
display:block
}
Notice I didn’t name the division something that screams I’m a spammer like “hidden”, “invisible”, “hide” or “HeyGoogleThisIsSpam.”
As Herman Sherman pointed out yesterday, “you can always pop that css file in your images directory … which so happens to be the same directory i always exclude in my robots.txt”
Posted in 
it seems to me that disallowing your css in your robots.txt is a bit obvious, and could itself be cause for reducing the relevence of a page. i am not sure if google or other search engines are this aware the practice, but rest assured they will become so.
i do not know how much the search engines parse css attributes, but it would seem plausible if they had access to the css file, they could identify links that are visibility: hidden or negative positions like top: -10000px.
a better solution may be to have your css served by a script which serves the search engine spiders(or anything that is not a popular browser) different css files. or maybe even serve bots a 404 not found on the css file.
with absolute positioning you have nothing to worry about if you want to place it somewhere visible but out of the way … altough getting it to work with a footer that isn’t always in the exact same position is a bit tough.
serve bots a 404 not found on the css file …oooh … wicked indeed.
you could just overlap an image over it as well.
oh yeah … a good trick too… use javascript to make a div id hidden. your css files will be clean.
can anyone recommend a tool to make the JS hard to read by an engine?
I wouldn’t bother obfuscating your js, if one day the bots learn to parse js they’ll have a parser built in so obfuscated code will be just like normal.
if they are doing lazy checking (looking for keywords within a js file) then just build up the hidden as a series of strings.
eg. s=”h”
s+=”i”
s+=”d”
s+=”d”
s+=”e”
s+=”n”
alert(”you wanted: “+s)
or for something a little more gray, make your links identical to the text arround it, no text decoration, no weighting.
[...] Interesting stuff from the world of Blackhat SEO…we all knew it but didn’t think to use it in part…I suppose [...]
[...] I was tipped off to Wikipedia Eats Google, through Steve Rubel’s Wikipedia all over Google. One his main proving points is this special query over at Google [* *] which is intresting, and this is the authors line of reasoning: Google is becoming increasingly prone to Wikipedia. This is because Google’s PageRanktm algorithm, the method by which it ranks search pages, inherently succumbs to the basic structure and social structure of wikis. The PageRank algorithm is most famously characterized as valuing links that are highly referred to by other people. It seems that is only part of the story. The PageRank algorithm values links to yourself more. That is, a website that has many pages and is densely inter-woven with links becomes a sort of PageRank machine. True, without other websites conferring a little bit of their PageRank onto it, that website will not have a high PageRank, but given enough of a small number of external links from mediocre websites pointing to your very large, densely interwoven website, your website will shoot up through the listings. However in true Freakonomics (no aff) style I’m going to say the author misses a big part of the equation. There have been many back and forth debates over whether linking out to unique, and authoritative web documents help your rankings (see Hilltop Analysis), however I can tell you I’m a firm believer in the concept, as are many other SEO’s. In fact many who practice the dark arts also link to authority websites. As a result of Google’s Sandbox filter I can tell you my site building skills have improved dramatically. I can crank out websites with hundreds of empty pages waiting for copy in a few days. Now if I were a true button pushing spammer I’d be filling these with scraped and uniquified content, and I’d certainly be linking to authority websites like Wikipedia, to make my site appear more “hub like” and rise in the rankings. [...]
I have a variation on the hidden text using CSS. I created an entry in my site CSS that has the position commands in it. This way you can make a very normal looking span statement and the text is invisable to human eyes but very visable to the search engines. I also put a lot of spaces before the postion statement to push it out of sight just in case someone downloads my CSSS file. I no longer use it but I still think it’s a very neat way to hide text.
.Arial8ptd {font-family: Arial, Helvetica, sans-serif; font-size: 8pt; color: #CCCCCC; position:absolute;top:0px;left:0px;width:200px;height:200px;visibility:hidden}
out bound loinks = no page rank reduction? i was looking at a few sites which shuold have a higher pr than they do, take for example google news. pr5, but maybe tahts because they dont havec any real content, just a few outbound links. anyone else can verify this?
[...] Myspace strips off the word Javascript, so he just split the word ‘JavaScript’ into two lines in his CSS. oojee covered something similar in the comment section of SEOblackhat here. [...]