SEOmoz released its latest version of the search ranking factors document that Rand Fishkin originally compiled a couple of years ago. He invited about 35 people to participate in the survey this time around and the new article, in my opinion, accurately reflects the SEO community consensus on many points.

Which is to say there is a fair amount of nonsense in the article. Like its predecessor, the search ranking factors article is an amalgamation of opinion, not a benchmark that sets any sort of standards for search engine optimization. In any community of thought, the consensus or majority agreement on issues follows a classic bell curve pattern when you chart points ranging from “I strongly agree” to “I strongly disagree”.

That means that your consensus map represents, at best, mediocre quality in whatever terms you want to consider. Or, to put a more positive spin on it, you could do much worse than to follow the advice given out by the majority of search engine optimizers. But you could equally do much better than to follow that same advice. It depends on which side of the curve your own choices fall.

I could nit-pick or applaud various points in the survey but it really doesn’t have anything to do with SEO theory. On the other hand, while we were glancing through the results at work earlier today, someone asked me a very interesting question that does fall under SEO theory.

Besides me, we have 3-4 experienced SEO technicians and 4 other specialists who provide support for various SEO projects. Those 4 other specialists are studying SEO principles and theory. So I do not exaggerate when I say we engage in theoretical research and discussion on a daily basis. But sometimes we just swap opinions and today I was asked what I think about internal links versus external links.

The SEOmoz survey offers some very bizarre classifications, such as “link popularity within the Site’s internal link structure”, “quality/relevance of links to external sites/pages”, “global link popularity of site”, “topical relevance of inbound links to site”, “link popularity of site in topical community”, and “global link popularity of linking site”.

So as I considered possible reasons for including such irrelevant classifications in a “search ranking factors” document, I was asked whether I feel internal links count as much, more, or less than external links.

It seems like a pretty good question, though it coud be asked many ways. The answer is that, in my opinion, it depends. What it does not depend upon, however, is how many links point to other sites. Nor does it depend on how many “global” links (what, exactly, is a global link?) point to a Web site.

Since the survey doesn’t define any of the unusual terms it uses (like “global link popularity”) we have to guess what these terms might mean — and did the respondents all use the same definitions or not? That question alone points out the unreliablity of the survey, but let’s assume for the sake of this discussion that “global link popularity” refers to all the links that point to all the pages on a site.

Who links to page A on a site has no impact on the value of a link from page B on that site if there is no chain of linkage from page A to page B. So links pointing out from page B are not going to pass on any value from page A’s inbound links. Hence, “global link popularity” has no impact on search engine results.

It may have an impact on trust.

Most SEOs today seem to be of the opinion that internal links don’t help boost your rankings for competitive queries. I’ve never understood that point of view, but time and again I have seen people say truly irrational things like, “Internal links help with uncompetitive queries but not with competitive queries”.

Really? Well, let me sell you some sand bars in Texas, because obviously you’ve got money to burn if you believe nonsense like that.

Links are links. Just because links are internal to a site doesn’t mean they should be trusted less than external links. In fact, your internal links are more likely to pass value in some form than any random external link. Neither you nor the search engines really knows where that external link has been. It needs to go through a bit of filtration before it can pass value.

But the search engines have never mentioned filtering internal links. No academic paper has ever mentioned filtering internal links. And while I have found many a spam directory doesn’t pass value to other sites, I (and millions of other surfers) have seen how they manage to get crawled.

If a crawler follows a link, the link passes at least some value.

So that means search engines will trust internal linkage much more easily than they will trust external linkage. Internal links possess no more innate value than external links, and they may pass less value than really good external links, but they’ll pretty much pass some value and that is more than what many external links pass these days.

Given a choice between a free external link from some obscure PR 9 Web site I’ve never heard of and 100 of my own internal links, I’ll take the 100 internal links every time without hesitation. I control those links. I know where they come from. I know where they point to.

In search engine optimization, trust begins at home. If you don’t trust your own content search engines won’t trust it either. And in that respect neither Yahoo! nor CNN nor any other Web site is going to pass more trust to your pages than your own pages.

SEOs get all wrapped up in (internal) PageRank. Well, I say PageRank ShmageRank. It has never had as much influence in search results as link anchor text and it’s unlikely to ever surpass link anchor text in value. Link anchor text is the search engines’ endorsement of spam. You can spam your way to the top of search results all day long with link anchor text. Search engines not only don’t care if you do that, they encourage you to do it.

And that is why SEOs wrongly believe that links are necessary to achieve high rankings in search results. After all, you’re not allowed to put “SEO theory SEO theory” on a page 500 times, but there are no rules against getting 1,000 other sites to link to you once each with the text “SEO theory”.

Either way, you get to add “SEO theory” to your content 1,000 times. When your document says “SEO theory” 1,000 times, you’d better believe it’s going to ring bells and hit the top of the search results for “SEO theory”. That’s just the way it works.

You can boost your relevance score by bolding SEO theory and by italicizing SEO theory. You can even put it in an H1 header:

SEO theory

Search engines like it when you put your keywords into your title tag, your page URL, and occasionally even your meta keywords tag. They like it when you spam them to death.

So what’s the problem with internal links? If search engines let you repeat keywords endlessly, why should they care about whether your links are internal or external? If anything, internal links should count more than external links because internal links are placed with editorial care. External links might just be spammed.

If your pages are stuck in the Google Supplemental Index, your outbound links won’t pass much value to other sites’ pages. Your anchor text will not be associated with those sites’ pages by Google. But the Supplemental Googlebot will still follow your links — at least enough to index your site.

Maybe Supplemental Googlebot will crawl from one supplemental site to another, too. I’ve never seen any Googler deny that would happen. Neither have I seen them say it would. All I know is that I have seen new sites show up with all or most of their pages in the Supplemental Results Index. Those pages didn’t all get into the Supplemental Results Index through external links.

But the whole question of whether internal links matter more, less, or about the same as external links presupposes that search engines weight linkage on the basis of location. The literature has never indicated this happens. No search engineer has, to my knowlege, ever implied or stated that internal links matter more or less than external links.

They do tell people to get links from other sites. But you need links from other sites to be crawled and indexed.

On the other hand, some spam analysis papers do talk about where links come from. They do distinguish between potentially spammy links and potentially good links. Of course, those papers are all concerned only with external links. I have never read a paper that proposed internal links should not be trusted.

Still, most SEOs spend a lot of time looking at backlink data. They seek out new external links. Now, don’t misunderstand me. I also go for external links. I know how to use them. I know how to pick them. But I put more faith in my internal links than I put in external links.

Backlink analysis is unrevealing in these matters. That’s not to say backlinks don’t help. Of course they do. You just don’t know which ones help.

Assuming you could identify all the backlinks Google knows about for any competitor’s domain, you won’t know which of those links pass value, you won’t know which of those links helped get the competitor crawled. And yet, knowing nothing about a competitor’s backlinks, most SEOs solemnly swear that those backlinks help those competitors achieve their high rankings in search results.

Maybe a page ranks well for a query simply and solely because it has 1,000 internal links pointing to it with relevant anchor text. Link anchor text is like sand. Put enough sand together, apply high pressure to it, and you create stone. Put enough link anchor text together, pass it to one page, and you create strong relevance for a query.

I’ve been able to adjust my rankings for competitive expressions simply by changing internal link anchor text. Of course, I have sites with lots of “global link popularity”. In fact, I know more about my backlinks than most SEOs know about their own backlinks. I created most of them.

Think about that.

Google Webmaster Central divides links into “external” and “internal” links. So now we know where all the (reported) links come from. But we still don’t know which links are not being reported and which links are not passing value. And I don’t mind telling you that many of my inbound links are transient. So even when Google tells me it found a link on a page only a few days ago, I can visit that page now and find no sign of a link.

Give me a link that passes value and I won’t care where it came from. It’s as simple as that.

The only real difference I see in internal links versus external links is that I am far more ready to depend on my own links than I am to depend on yours. I don’t know if your links will pass value. What I do know is that, if my links don’t pass value today, they will. I won’t settle for anything less.

The only good link is the link that passes value, and where it comes from is completely irrelevant in that respect.