| Relevance and Reciprocal Linking |
|
|
|
Written by Michael Martinez January 2, 2007 Reciprocal Linking from an SEO PerspectiveWith all the discussion about reciprocal linking lately, I thought I would add something to LED's comments.Reciprocation in itself is not a bad thing. The search engines see it all the time, and earlier this year a technical paper titled "Site Level Noise Removal for Search Engines" (this is a .PDF file) provided a simple algorithm any search engine could be using to identify all reciprocal links between two hosts (which I'll loosely define as either a distinct domain or sub-domain). What we call reciprocation they call "site level mutual reinforcement relations". My guess is that Google has probably been using this method to identify reciprocal link patterns for at least two years. Maybe longer. It WILL identify virtually all reciprocal HTML links between two sites. Nonetheless, though I no longer participate in organized reciprocation, the sites in my network openly reciprocate with each other without penalty. In ways they have not fully disclosed, one or more of the major search engines attempt to filter out some reciprocal links. The question is, which links don't help? We can make some educated guesses. For example, a lot of people are now buzzing about "relevant links", though few people have tried to explain what they think a relevant link is. Many people imply they only exchange links within their verticals or industries. I don't believe that is what the search engines care about. They are looking at on-page content first when they determine relevance for queries, so why not look at on-page content when determining relevance between linked pages? But relevance doesn't always matter. CNN can link out to an obscure Web site as an aside in an article and that site gets some juice. Why? Or, rather, why don't all sites pass juice like CNN? Because some sites have been caught linking out in ways search engines disapprove of. Or for other reasons the links don't pass value. Maybe they are using "nofollow". So, you can get links from relevant content that don't pass juice. If search engine optimization is the only reason you're trying to get links, you're doomed because you have no way of knowing in advance which links will help you. Still, a link that doesn't pass value in a search engine may still send you traffic. And such links may still help get your site crawled. It's not all about search engine rankings. But it shouldn't all be about search engines anyway. Besides relevance and value, there are the issues of trust. "nofollow" is the Webmaster's way of saying, "I don't trust the site I am linking to enough to recommend it". But search engines may not trust a page for other reasons. For example, earlier this year Matt Cutts wrote on his blog that "sites that fit 'no pages in Bigdaddy' criteria were sites where our algorithms had very low trust in the inlinks or the outlinks of that site." Who you link to helps the search engines determine whether you're linking out naturally or just trying to build up your link popularity. I have seen evidence that Google has been evaluating outbound links for years. I believe that with the Bigdaddy update they upgraded their evaluation of outbound links considerably. A natural concern many people have is whether they may get into trouble for link arrangements that make good business sense regardless of the possible influence such links may have on search engines. In my new blog, I wrote about what I call Semi-natural linkage earlier this month. I think the search engineers are aware of these types of relationships and I don't believe they want to devalue them. I see no evidence of intentional, across-the-board devaluation of trust in Semi-natural linkage. Well, all that is to say that I don't think reciprocal linking is dead. And, strangely enough, the Google blog post that started the whole fuss didn't even mention reciprocal links. I think that organized reciprocal linkage has been under very close scrutiny for several years. I think that sites which put up links pages probably don't pass much value -- if any -- through those pages. But I think that natural, business sensible reciprocation is acepted because it's neither abusive nor designed to game search engines. I think that most reciprocal link efforts made today are in vain because they follow the traditional model of adding a link to your reciprocal links page. If you're going to seek reciprocal links, you should seek links that are naturally compelling for people who would want to visit your site. Whether you seek them out yourself or sign up for a service should not matter. But if you sell golf clubs and you reciprocate with a ring tones affiliate site, think long and hard before you stand up at an SEO conference and ask Matt Cutts why your rankings suck. Michael Martinez seo-theory.blogspot.com Comments (3)
R. Deisler
said:
|
|||||||
| Interesting points, sensible post. But I sense a contradiction which makes me wonder about relevance even more in this discussion. You said first, "Many people imply they only exchange links within their verticals or industries. I don't believe that is what the search engines care about. They are looking at on-page content first when they determine relevance for queries, so why not look at on-page content when determining relevance between linked pages?" Fair enough. But then why do you end with this contrary idea? "Whether you seek them out yourself or sign up for a service should not matter. But if you sell golf clubs and you reciprocate with a ring tones affiliate site, think long and hard before you stand up at an SEO conference and ask Matt Cutts why your rankings suck." Going by the logic from your first point, if my golf club site had a page about cool ring tones Tiger Woods uses and I link to an affiliate or lead gen page... get my point? If it's on-page content that matters then relevance doesn't ... if it's not then it does. So which is it? |
Jeremy Kingston
said:
| if you watch the posts that Martinez makes in the webmastering community, you will see a trend where he is almost constantly contradicts himself. This is probably because his comments are self-serving in order to drum up business. If he had actual clients, he would spend more time servicing them instead of contradicting himself in forums throughout the web on what seems to be an almost daily basis. Google is working to determine if sites link in heavy volume to junk sites or if sites that rely heaving on link exchange do so within ethical limits. That took only a few sentences to explain. Why does it take martinez multiple contradicting and confusing paragraphs to muddy up this simple concept? Be very wary of self proclaimed seo "experts" who complicate matters and contradict themselves. Think about it.. why do folks like martinez post in forums so often such as this? answer: no clients and nothing else to do. |
Michael Martinez
said:
| You know, taking things out of context makes it easy to pretend that someone has made a contradiction, but that's pretty superficial and easy to spot. The first comment addresses the reciprocal link myth. The second comment addresses the reliance upon reciprocation with spam sites. Hence, there is no contradiction since my remarks were quoted out of context. It would be far better for people to beware of those critics whose only complaints are the fabrications they place in other people's mouths. |
Write comment
