I’ve been on an anti-Google tear recently (privately, Facebook, nothing really public).  The more time that goes by, and the more the company is observed, the more they are observed go against their self-proclaimed “don’t be evil” creed.

An example from about 7 months ago is a blog post by Matt Cutts, the guy every SEO-obsessive type seems to listen to the most:


In the post, he describes the change in algorithm Google applied some time in 2008 in regards to rel=”nofollow” links.  I’ll get to why it’s all a big cluster fuck in a moment but first some background on me:

I’m the kind of guy who doesn’t spend a lot of time on “SEO” concerns or jumping through every hoop those types jump through.  I focus on the creation of good ideas, good content, good value web-driven resources and the rest SHOULD sort itself out.

The problem is that the kind of content I maintain is followed by a lot of people who link back through blogs, fora, social sites, etc, and when Google announced the deployment of rel=”nofollow” a year later a lot of web frameworks, blog & forum software and hosted service sites deployed rel=”nofollow” on all user-submitted content and, further, webmasters all over went a step beyond and applied rel=”nofollow” to almost everything.  The impression was that it was a way to salvage or enhance “link juice” when the main think (allegedly) Google was trying to do with it was reduce the amount of spam links they inevitably followed.  It was a network resource saver for them as well as a way for them to reduce the processing load of figuring out the rankings of billions of pages.

The result when the dust settled was that people who don’t concern themselves with SEO and just simply manage good content got hurt.  Organic search got affected and it forced those types like myself to have to pay attention to SEO.  My theory is that Google’s strategy was to force more web sites to play the SEO game and thereby create an army of webmasters who jump every one of their hoops in order to shape the web’s reliance on them in search dominance.

Ho nofollow originally worked: If a page of rank 10 had 10 links on it, a rank of 1 was passed down the pipe to each of those links.  If 5 of the links were set to nofollow, the rank would be divided among the remaining links (2 each) which then causes SEO people all around to “hoard” link juice by using nofollow on even more links as a way to “shape” page ranks throughout their sites.  Well, as Matt described in his blog post, the new method as of 2008 now causes rank to leak even to those nofollow links, causing the remaining links without nofollow to have less “link juice”.

The end result of this is that they made using nofollow necessary and now they’ve made it less-than-useless.  In fact, it’s now downright detrimental.  What they’ve FURTHER done to dodge people simply removing nofollow is to actually count any links AGAINST sites if those links are otherwise pointing to stuff their engine thinks is spam or bad content.  Technically, this should NOT affect me, since I don’t allow active links in the first place rather than allow links to sites I may not otherwise perceive as valuable.  But, this DOES affect me because it further degrades the value of links coming INTO my sites because I’m pretty sure a lot of such sites that link to my resources play the “SEO game” too much and therefore have degraded the value of links throughout their site due to following the nofollow mantra too closely.

I have NOTHING to do with sites who link to me but having links come back to me is SUPPOSED to be a sign of authority.  But, the way Google has tiered this deployment, first the organic click value was reduced to rel=”nofollow” being deployed and now it will degrade even more because the remaining links are reduced in passing rank.

Cut in half once, then cut in half again.

Do you see what’s wrong with this picture?

But, I have a solution, one that I’m sure Google will eventually work around.  If they want people to play the SEO game, then I’ll be on the side of people who take it upon themselves to “game” them.  They want to retain value in their search engine, and the way to do that is make people jump through hoops to play their game.  If they want people to play their game, then I’ll be on the side of out-gaming them.

It’s a simple matter of math.

If Google passes rank based on the number of links on a page, still counts nofollow links to do their division but no longer give back that diluted rank to the links on the page without nofollow, then the way around it is to NOT remove rel=”nofollow” on those links unless you know for sure that those links don’t point to something Google considers spam.  Since you don’t know how to figure that out (who does?) then your option isn’t in removing rel=”nofollow”.  Doing so won’t HELP you and it may even HURT you.  But, if you get value from those links (like, perceived value of visitors of sending them somewhere valuable, or they are paid links or point to other web sites you run) but no need for passing rank, then you need a solution that allows you to STILL link to those resources without those resources counting AGAINST you and without those resources diluting the other links on your pages.

I don’t like this but the solution in those cases, where possible, is to re-code those links with JavaScript.  95% or more of people now have JavaScript enabled.  For those without JavaScript enabled, a simple noscript containing a URL that isn’t hyperlinked will suffice.  That addresses usability issues in a reasonable way.

On my own sites, I’ve simply dodged the allowance of people to leave links in the first place.  I’ve also dodged the use of rel=”nofollow” except in the rare circumstances where it made the most sense.  I can only do things that improve how I pass rank, simply because my focus all along has been to maintain high value resources, something Google keeps telling us is what they want (yeah right).

However, since I am likely affected by this whole nofollow fiasco in the context of outside sites linking into to mine, I will give all you SEO-obsessed guys this little solution:

<script language=”text/javascript”><!–
document.write(’< href=”http://outsidesite/”>Great Resource</a>’);
–><noscript>Great Resource: http://outsidesite/</noscript>

Now, the (allegedly intended) spirit of rel=”nofollow” is, er, followed and the remaining active links on your site get to retain more “link juice” than they were left with based on the most recent change.  Those of you who want to render via the DOM can do so, the above is just the most basic example to start with.

Now a page with a rank of 10 with 5 direct links and 5 nofollow links will have the rank divided among the direct links. Further, it gives you a means to dodge some future spanking they might deploy if they decided that they’ll degrade the rank of your pages/site based on even links you’ve designated nofollow.  It dodges the issue entirely, and allows you to go back to “re-shaping” and not cause grievance to the sites you do link to directly because the rank you pass them is back to being “shape-able”.

This will likely get fucked in a year or so when Google is able to start following JavaScript links and treat them as raw HTML, but that’s just another hoop for Uncle Google that you’ll have to jump through, isn’t it?

bookmark & share:
  • Digg
  • del.icio.us
  • Reddit
  • StumbleUpon
  • Google
  • Facebook
  • YahooMyWeb
  • BlinkList
  • blogmarks
  • Fark
  • Furl
  • Live
  • Propeller
  • Pownce
  • Bumpzee
  • BlogMemes
  • Slashdot
  • Spurl
  • Technorati