You measured PR and found a strong correlation. Certainly, correlation is not causation, so you could be right — but where is the evidence? I wonder why you would do such a study, only to dismiss the results as a byproduct of something you didn’t include in the study. On your own blog you say:

“When people make statements but fail to back them up with evidence, BEWARE!”

My point is not to say that PR is a primary factor in ranking — but that it is a good indication of how important Google thinks the page is. The same page will rank differently for each keyword/keyword phrase — there is no single overall ‘rank’ for a page. There is a single overall PR for a page, so it is a useful guide to how well you are doing in getting Google to consider your site important.

PR is also a useful factor to consider when working on getting incoming links. It is not the only factor — but given two sites of comparable authority, getting a PR0 link on one would not be as good as getting a PR5 link on the other. If getting those links requires effort on my part, I’m better off making the effort to get the PR5 link.

So PR does matter, even if it is a byproduct of good ranking rather than the cause (though your study does not show any evidence of that, you simply assert it without proof).

That 1-10 list would look precisely the way it looks now if PageRank doesn’t matter, because PageRank is a byproduct of having a lot of links to your site, and a lot of links mean good rankings. So naturally, since lots of links result in both a higher PageRank AND good rankings, it would APPEAR that PageRank has a big effect. The reality is, though, that it’s the links causing the ranking, and not the PageRank.

You see, I can go get a single link from a PR8 site, and in the next PageRank update my site would have a PR6 or PR7, but it wouldn’t rank for anything worthwhile because it only has one link to it! On the other hand, if I got thousands of links into my site, which resulted in a PR6 or 7, then I would easily be able to rank for substantial keywords. It’s not the PageRank that gets the site ranked, but the links. But since higher PageRank is a side-effect of having a lot of links to the site, it APPEARS PageRank matters much more than it does.

If PageRank was NOT the byproduct of links to the site, then you would be correct in saying that there would be an even distribution across the ranking points if PageRank didn’t matter. But since PageRank is the byproduct of links, and links are what cause the rankings, it will not be an even distribution at all — sites ranking higher will TEND to have a higher PageRank because more links generally means more PageRank. But the fact that this is NOT the case 1/3 of the time demonstrates that it’s not the PageRank causing the ranking. Further digging demonstrates that it’s the links to the page, and not the PageRank of the page, that matters.

I wish PR was a primary factor in ranking. If it was I would run out and get links from high PageRank sites and get a bunch of brand new sites with very few links ranked like crazy.

Thanks for the post.

Jonathan Leger