How Is Google Ranking Knol Articles?
Posted: Tue Aug 12, 2008 12:25 am
Some people noticed Knol articles are already sometimes ranking very well in Google results, even though Google promised Knol articles wouldn’t get any artificial boost. Until further evidence comes in I don’t think we’ve much reason to distrust Google’s statement, and yet, the site does get a major boost all the same simply because it’s in the vicinity of the superbly ranking network of Google websites (not yet on the PageRank 10 homepage, but for instance on the PageRank 8 – recently 9, AFAIK – blog, on PageRank 9 Google Scholar, and perhaps now or soon on other Google properties... not to mention all the other press backlinks Knol gets as it’s a Google project). But don’t expect any guaranteed high rankings; an article I started on the competitive subject of search engine optimization did not jump into the top 10 for that phrase.
At the same time, Google can’t be neutral ranking its own properties even if they don’t directly, manually mess with individual rankings – that’s because they hire the testers which evaluate their search result algorithms, and we have reason to believe these tests in turn can shape which directions Google’s algorithms take. It would be hard for any search company to not be trapped in this conflict. But in the future, what could be a more direct skewing with results would be when Google tries to format Knol results differently from the rest; like by adding eye-catching star icons near the snippet to show how well an article rates. Or just imagine a special “onebox” listing Knol articles amidst organic results; Google already does this with Google News content, for instance, but they don’t host all that content themselves (though they do host news from some agencies).
Additionally, Google is already somewhat messing with link juice by counting most of the incoming links to Knol articles, but completely disregarding the outgoing links (on the articles I checked, at least), thanks to the use of nofollow. (Not that they’re alone in that; Wikipedia does the same.)
Incidentally, before Knol’s release a common attack against Google was that Wikipedia articles rank way to high way too often. This was something I never found to be true, which is of course a matter of web site taste and preference. Often when I want to find a good introduction to a given topic I look at nothing but the Wikipedia article at first, and then go from there to perhaps find comparative sources. Entering search query wikipedia or similar yields Wikipedia often enough, and even entering just search query does, quite frequently. In a way, that might be because often a given Wikipedia article is the closest thing to a “mass consensus” kind of article you will find online today. We’ll see how Knol compares with this, because many or perhaps most of the Knol articles are anything but mass consensus... Knol is very much single-author-driven, and even if you pick the default publishing mode (which allows revisions from others, but each will be moderated before a potential go-live) the name of the original author will still appear visibly on top,
At the same time, Google can’t be neutral ranking its own properties even if they don’t directly, manually mess with individual rankings – that’s because they hire the testers which evaluate their search result algorithms, and we have reason to believe these tests in turn can shape which directions Google’s algorithms take. It would be hard for any search company to not be trapped in this conflict. But in the future, what could be a more direct skewing with results would be when Google tries to format Knol results differently from the rest; like by adding eye-catching star icons near the snippet to show how well an article rates. Or just imagine a special “onebox” listing Knol articles amidst organic results; Google already does this with Google News content, for instance, but they don’t host all that content themselves (though they do host news from some agencies).
Additionally, Google is already somewhat messing with link juice by counting most of the incoming links to Knol articles, but completely disregarding the outgoing links (on the articles I checked, at least), thanks to the use of nofollow. (Not that they’re alone in that; Wikipedia does the same.)
Incidentally, before Knol’s release a common attack against Google was that Wikipedia articles rank way to high way too often. This was something I never found to be true, which is of course a matter of web site taste and preference. Often when I want to find a good introduction to a given topic I look at nothing but the Wikipedia article at first, and then go from there to perhaps find comparative sources. Entering search query wikipedia or similar yields Wikipedia often enough, and even entering just search query does, quite frequently. In a way, that might be because often a given Wikipedia article is the closest thing to a “mass consensus” kind of article you will find online today. We’ll see how Knol compares with this, because many or perhaps most of the Knol articles are anything but mass consensus... Knol is very much single-author-driven, and even if you pick the default publishing mode (which allows revisions from others, but each will be moderated before a potential go-live) the name of the original author will still appear visibly on top,