If you missed the original revelation of Google Knol, you might want to check our coverage of it. That article includes a definition of what a knol is, a close look at the sample knol Google was showing off at the time, and some analysis of the likely effects on SEO. I noted at the time that knols were probably the ultimate linkbait, and that if you’re an expert in your field, this was your chance to write the introductory article that everyone would turn to when they want to know about whatever it is that you do. I also noted that links in a knol might not be nofollow (unlike Wikipedia links, which are nofollowed automatically), so they certainly had potential to earn you some traffic. As it turns out, I was wrong about that last point – but links to your knol from the Google Knol home page aren’t nofollowed, and if your knol is featured on that page, it could mean lots of traffic.
I also mentioned some potential minuses to Google Knol. One of them was user authentication; how could Google be absolutely sure that a person writing a knol really is who they say they are? The second issue was spam, in that there didn’t seem to be anything in place to keep a spammer from writing a knol and then using his or her friends to push it to a high ranking.
I shouldn’t say “I told you so” at this point, even though I’ve seen some reports online that Google’s user authentication isn’t working. I didn’t predict that spammers might steal someone else’s content, post it to a knol as their own, and then manage to score higher in the SERPs than the original content. Has this happened yet? Not quite, but Aaron Wall’s experience, which I’ll discuss shortly, seemed to prove that it is at least possible.
By the way, if you’re looking for an article on how to write a knol, this isn’t it. I’ll be analyzing, not instructing. If there is interest, I might write a knol and cover the process of doing so in a future article. If you’re looking for something a little sooner than that, Elinor Mills at CNet did a piece on pit bulls for Google Knol and describes the process of writing a knol.
Everybody except Google is describing Google Knol as the Wikipedia killer. While Wikipedia authors are anonymous, knol authors are verified and accountable. Medical articles are written by real doctors. Take the article titled “Migraine: Mechanisms and Management.” The author, Richard Kraig, is a neurologist and neuroscientist at the University of Chicago Medical Center. The first thing that shows up on a search of his name is his bio with the medical center.
On top of this, only the author can edit his or her own knols. Others can propose edits, but the author has full control over the knol. Google doesn’t even edit the knols, but authors may, at their discretion, allow Google to include ads in their knol. The search engine will share revenue with authors who allow the display of ads.
Right off the bat, that raises “conflict of interest” flags. If Google is getting income from knols, wouldn’t it be tempted to rank these higher than other sites? The search engine claims that it won’t do this, because, after all, it would be evil. But Danny Sullivan noticed something very peculiar concerning knols and their search engine ranking. A day after Google Knol went into public beta, Sullivan did searches for the titles of 30 knols on the Google Knol home page. Before I tell you the results, keep in mind that one day should be too early to be ranking in the SERPs, especially for competitive terms targeted by long-standing web sites.
The results? One-third of those 30 knols hit the first page of Google’s SERPs for their title. As Sullivan put it, “It’s proof that being in Knol is NOT an automatic ride to the top of the search results. But then again, knowing that 33% of your stuff will rank within a day is a pretty good track record.” Sullivan further notes that back links might be helping knol rankings more than one can easily determine (remember, Google doesn’t show all back links).
It appears that Google trusts knols more than they trust content on the web at large. Reading between the lines, creating more trustworthy content seems to have been Google’s intention all along when they created Google Knol. But that doesn’t mean that the content actually IS more trustworthy than what’s already out there, as Aaron Wall discovered.
Aaron Wall noticed the results of Danny Sullivan’s test, and some other peculiar items. He decided another test of Knol was in order. He wanted to see how Google would handle a knol if it was duplicate content of something promoted somewhere else online. And he had some ready made content to do that with – his own guide to learning SEO, which was also syndicated to Business.com.
The Business.com page with Wall’s content has a PageRank of 5. But after Wall built his knol and searched for a string that was on both the Business.com page and the knol, guess which one ranks higher? If you guessed that the knol did, then not only are you right, but you probably understand why Wall is so concerned.
“Some might call this the Query Deserves Freshness algorithm, but one might equally decide to call it the copyright work deserves to be stolen algorithm,” Wall notes (emphasis in original). Google clearly knows that the knol is duplicate content, but they still ranked it higher than the much older content from a reputable site.
So take this one step further. Say someone writes a knol using your content and manages to get a few decent links to it. Even if your site is an authority site, you could find yourself slipping down the SERPs. That could cost you traffic, and by extension, money. If you make your living by publishing copyrighted content to the Internet, it’s enough to make your blood boil. Sure, Google’s approach might take down Wikipedia as an authority site – but if Wall is right, it could end up tarring other authority sites with the same brush.
So we know that knols can promote duplicate content. What about defamatory content? In Elinor Mills’ CNet article on writing a knol, she mentions that Wikipedia co-founder Jimmy Wales’ ex-girlfriend Rachel Marsden wrote a knol about him in which she described Wikipedia as an “online libel board” that “any loser can use to smear people who are more successful than them.” Google said in response to Mills’ questions about that knol that they aren’t editors, and that there are community tools in place to help flag items that cross legal lines. “In addition, because knols are attached to verified author names, we think that the structure of Knol will actually provide something of a disincentive to defamatory or other harmful content,” the Google spokesperson said.
I think something must be working here, because when I searched for that specific knol using several different keywords, I couldn’t find it. One imagines the community flagged it enough to bring it to the attention of someone at Google who realized this could cause issues. Google may be a search engine, not an editor, but all of a sudden it’s faced with the same concerns of a content company.
Which brings us to the most dangerous point of all. Google isn’t just a search company any more. As Jason Calacanis observed, it’s now a content company. He doesn’t buy protestations from David Eun, Google’s head of partnership, that Google has no aspirations to become a media company. “We don’t produce content. In fact, we see ourselves as a platform for our partners that do,” Eun said.
According to Calacanis, though, Google is splitting hairs that don’t even exist. He made a list of five important things that a content publisher does, and showed that Google Knol does all of these things. Here’s his list:
- Hire writers. Google can be loosely interpreted to do this on contingency.
- Distribute their work. Calacanis maintains that Google distributes their work in its own search results – which it arguably does, given the points I made earlier about Google favoring its own knols over other authority content.
- Sell advertising against the content. As I mentioned, that’s an option that any author can opt into.
- Pay the writers. Google does that via AdSense split. Just because some writers can opt out of payment doesn’t mean that others aren’t getting paid. After all, even newspapers have letters to the editor, which they publish for free.
- Build a library of that work for future monetization. Yes, Google is clearly building a knol archive.
Just because Google itself doesn’t own the content doesn’t mean that it is not in the content business. Given that a lot of publishers get a huge chunk of their traffic from Google, and that Google is ranking many knols higher than similar content on authority sites…well, you can do the math. Calacanis notes that Google getting into the content business, when they already dominate the search business, looks like a play for a monopoly.
“This feels exactly like what Microsoft did to its application vendors,” he explained. “Microsoft convinced folks to build WordStar, WordPerfect, Lotus 1-2-3, and Quattro Pro for their operating system. They grew that business together until the point that Microsoft had massive market-share in operating systems – then Microsoft pulled the rug out from under the 3rd party application vendors.”
So is Google turning into a monopoly? It could happen. It seems like they’re trying to own more of a user’s attention, and that’s a valuable commodity when you’re selling ads. This situation needs to be watched very closely. And if you don’t have a knol yet, you might want to write one soon.