Best SEO Blog|Launch your site.

Two years ago I wrote The Relevant Link Myth on SEO Theory. The point of the article could probably be summed up by this paragraph:

The idea that a link may carry greater weight from a page that is devoted to the same topic as the destination page is another bit of nonsense that SEOs have adopted. How many of you would turn down a link from the front page of Yahoo!? I see no hands. Apparently, relevance isn’t all that it has been cooked up to be after all.

Today Rand Fishkin wrote an interesting post on SEOmoz, asking if it’s time to revisit themed links. I voted in the poll attached to the article and I picked “Yes. Theme-ing of external links is an important relevance/ranking element.”

Further on in the comments, I agreed with Russ Jones of Virante, who said: “Getting a themed link is not enough. Getting a link from a site that ranks for your primary term is.”

In my reply to Russ, I wrote: “I agree with Russ. I feel there is some sort of credibility aspect to Google’s link analysis. They seem to know when a link is off-topic and on-topic. Probably that’s a result of their pursuit of paid links.”

So, it’s logical to ask — am I trying to have it both ways or have I changed my mind?

A lot can happen in two years and I think you have to look at the circumstances behind the issue. Technically, what I wrote two years ago is still valid from the theoretical perspective. A link is a link is a link, as Jill Whalen likes to say. But not all links are the same, as most of us now recognize. Google has been trying to prove that PageRank is really a good idea (it remains the most stupid idea in search engine technology history).

We in the Web publishing community have to live with PageRank for as long as we care about Google, and for as long as Google cares enough about PageRank to try to make it work. That means that Google is doing everything it can to qualify links. Google wants to prevent links (that Google doesn’t like) from passing value and it wants to reward links (that Google likes) by allowing them to pass value.

You and I have no way of knowing which links work and which links are a waste of time. Unless you’re a link spammer with a horde of low-budget link placers or scripts running across a network of IP addresses, you cannot afford to rely upon the shot gun effect. You have to build your links with as much precision as possible.

That means you need to do a better job of qualifying where you get your links from than ever before because if you’re relying on links to build your search engine rankings (you should not be but probably are) then you need links that pass value.

The safest approach to precision link building is to find credible linking resources — and that means they should be topically related to whatever they link to. You don’t necessarily have to get a horse page to link to your horse page, but you should get your horse links from copy that is relevant to your horse copy.

The search engines are not perfect. They are mechanical, algorithmic, and lack the subtlety of human emotion, judgment, and experience. Nonetheless, they are improving their abilities to associate content with content. The search engines work with synonyms, homynyms, and topics in a growing number of ways. You can adjust your relevance guage a little bit to allow for some non-keyword specific placements.

I just feel that you really want to focus on credible linking resources. Following the old rules of thumb about relevant linking, themed linking will set you on the path toward working with credible linking resources. They won’t take you all the way, but you need to start somewhere and the relevant link myth is an easy starting point for many people.

I still say that Google would reward a link placed on the front page of Yahoo!, CNN, or WhiteHouse.gov because, frankly, such links would really have to be earned. They would be EXTREMELY credible. The anchor text would make them relevant as always but in today’s search engine environment I don’t think relevance is as critical as credibility.

Call it algorithmic credibility but we’re really being forced to look for links from credible resources in credible placements. Maybe last year’s quick link drop methods still work but why bet on that approach in the long run?

Now would be a good time to adjust your selection criteria for link placements.