News Source: http://searchengineland.com
Published: Jul 24, 2008 - 02:37 pm
Story Found By: northrock 3708 Days ago
Category: Vertical Search
70 Comments
Sorry this site requires JavaScript to be enabled in your browser. See the following guide on How to enable JavaScript in Internet Explorer, Netscape, Firefox and Safari. Alternatively you may be blocking JavaScript with an advert-related or developer plugin. Please check your browser plugins.
News Source: http://searchengineland.com
Published: Jul 24, 2008 - 02:37 pm
Story Found By: northrock 3708 Days ago
Category: Vertical Search
70 Comments
Upcoming Conferences
Search Engine Land produces SMX, the Search Marketing Expo conference series. SMX events deliver the most comprehensive educational and networking experiences - whether you're just starting in search marketing or you're a seasoned expert.
Join us at an upcoming SMX event:
Latest News From Marketing Land
Upcoming Webcasts
Learn more about search marketing with our free online webcasts and webinars from our sister site, Digital Marketing Depot. Upcoming online events include:
- Google Analytics 360: A Buyer’s Guide - Sep 27
- Google, the EU, and Comparative Shopping Services: Minefields (and Opportunity) Everywhere! - Sep 6
- Using Voice Search to Transform Your Local Search Strategy - Sep 11
- Protect Your Brand: The best defense is a good offense - Sep 20
- You Got The Click! Now What? - Sep 18
Research:
- Enterprise Customer Data Platforms: A Marketer’s Guide
- Account-Based Marketing Tools: A Marketer’s Guide
- B2B Marketing Automation Platforms: A Marketer’s Guide
- Integration Platform as a Service (iPaaS): A Marketer’s Guide
- Enterprise SEO Platforms: A Marketer’s Guide
Whitepapers:
- 7 Inbound Marketing Service Ideas to Grow Your Agency
- PPC Bid Management Guide
- The Marketer’s Guide to Driving ROI From Facebook Advertising
- Collaborative Customer Interactions Management
- How To Create a Data-Driven Dynasty
Latest News From Search Engine Land
- Amazon now 3rd biggest digital ad seller in US
- Wanted: Session proposals for SMX West
- Google: AMP to use ‘consensus-seeking’ governance model
- SearchCap: Bing drops URL submission tool, Ask the SMXperts, Google Maps, link building & more
- Bing Ads gives advertisers machine learning-powered insights
- - See more news at Search Engine Land -
Comments
Very interesting post and a little worrisome, what happens when google has 9 other sites and suddenly the only thing ranking in the top 10 are google properties...
It’s the only way they can ever hope to become as popular an outlet as Wikipedia. Not that they’re trying, right. ;-)
The best part about this whole Knol business is the Dutch steamcleaning equipment company that owns www.knol.com and apparently are not selling. :)
^^i feel bad for that company, you know there bandwidth bill has def. gone up.
Danny, aren’t the results of this experiment affected by the freshness effect? That is, as you know, fresh pages rank better for a time after they’re created.
I think Google made a poor choice naming this "Knol"I just searched for "Knol" in Google, and the KNOL stock ticker appears at the top, then news results for Knol (this could be skewed for relevance because of Google announcing it yadda yadda thinks its important and then the news results will appear lower again), then Google Blog, then Wikipedia’s entry for Knol, and then, finally Knol!Not the best user experience for those looking for the site.Also, when you type in "Google shopping" or "googe blogs" to get the sub-search engines, they are also not the top results.I’m just sayin...
The Google Knol I created just yesterday entitled ’How to Read the Russian Alphabet in 75 Minutes’ is already appearing on Page 1 of a Google Search for "read russian" with 12,300,000 results following. I have updated the title within Knol from the one showing in the Google index, but that’s still a pretty impressive result for content that I created and put online only yesterday.
So many of my clients pages reside in 3rd for their optimized keyword search term. (Brands using term within URL or company name #1, Wikipedia #2, client #3. Now i’ve gotta explain why they are now going to be #4...
Exact match shows only 57 results for "firefox plugins for seo" and 119.000 for "read russian". Nice, but not really shocking in terms of ranking, is it?
@bertifuel - Not shocking, I agree, but it’s the speed of ranking that I was talking about. The fact that one can CHANGE the titles and meta-data in Google Knol at any time, but still have the content served up in google’s search results when the description and URl has changed is also quite interesting.
Seth, query deserved freshness could indeed be a factor. Like I said, there are lots of caveats that you have to consider. Still, it felt like more than that.Bertifuel, yes, good point, the fact so few pages have that exact phrase can help explain why this particular one might be doing better.
The use of robots nofollow tag in the source code and nofollowed links on the page is interesting to see since it is a Google property. No problem ranking fast, the links are followed and seem valued. Kinda goes against the idea of nofollow.
I would think stuff like this leads to the demise of Google’s trustworthiness. If nobody trusts Google, they’ll start jumping off the ship. Google is pretty impervious at this point, but not so much so that they shouldn’t have SOME fear people will start going elsewhere.
On the flip side, I’ve had a few knols up for three days now, and they don’t show up at all in search results. I tested a few other recent submissions (none from the front page) and it’s the same deal...they don’t show up. I don’t know if Google simply isn’t spidering its own Knol site, or has put in some sort of lag (if so, it’s a long one!).One possibility is that they are withholding indexing Knol articles that have signed on to Adsense (as I have), until they go through some sort of clearance process that Google says could take up to two weeks. (I’m not new to Adsense...I’ve had my account for several years).If you want to take a look for yourself, to see what I mean, here are two of my knols:http://knol.google.com/k/david-sarokin/online-newspaper-archives/l9cm7v116zcn/5#Online Newspaper Archivesandhttp://knol.google.com/k/david-sarokin/earning-money-on-the-internet/l9cm7v116zcn/3#Earning Money on the Internetneither of which show up in any search results. Search on [ sarokin knol ] and you’ll see what I mean. It’s interesting that Danny’s article rose to the top of the search page like a rocket. Other folks can’t even get an appearance at all. I’d love to hear folks thoughts as to what they think is going on.Cheers, all.David
I left a comment on Dare’s blog urging caution before people jump to conclusions: http://www.25hoursaday.com/weblog/CommentView.aspx?guid=7fb38fd2-a79d-41fd-823e-e87c48c00543#commentstart<div></div><div>Apologies if there are any typos. I’ve been up late trying out cuil.com. :)</div>
Using SERPs as proof or even a reference to claims about Google algos is sloppy. Danny made the same mistake in his PageRank post, for example, when he claimed a TBPR 0 page outranked a TBPR 10 page so PageRank doesn’t matter, which would be true only if PageRank is the only ranking factor (and TBPR accuraltely reflects PageRank). That example doesn’t prove PageRank is irrelevant to rankings; it only proves PageRank (provided TBPR is a good indicator of internal PageRank - which we can’t assume) is not the only ranking factor.The recent second anchor link text also highlighted our tendency to assume we can isolate a ranking factor and draw conclusions about its effects. If there are other unknown factors effecting results, any conclusions you draw based on that assumption won’t be reliable.
Been trying cuil.com too - interesting presentation on results, and can see Google testing some inspiration with that - but aside from that, cuil doesn’t seem to have much to give. Nothing to give Google nightmares over. :)
My debunking post:Google Knol Ranking - Google NOT favoring Knol (the way you think ...)
seth your "debunking" is basically your opinion with no backup data whatsoever.
@halfdeck, Using SERPs as proof or even a reference to claims about Google algos is sloppy. "Danny made the same mistake in his PageRank post, for example, when he claimed a TBPR 0 page outranked a TBPR 10 page so PageRank doesn’t matter."What I’ve said is that PageRank is not the primary factor for why a page ranks well. Sometimes people think the page with the highest PageRank wins. That’s not the case. PageRank is relevant, but it’s one of many factors.@sethfinkelstein, to be clear, I didn’t say Google is overtly favoring Knol. Of course that would be stupid. I was saying that the site seems to have gained authority rapidly. And as an authority site, content within it -- in my opinion -- might rank better. That gives Google an inherent confiict just as they have with other sites like YouTube and Google Book Search. It’s a difficult path to walk, because as I said in my article, there are good reasons why they need to host content on YouTube and Book Search. There aren’t good reasons why they need to run Knol (someone else could do it, and there are alternatives).
"Sometimes people think the page with the highest PageRank wins."I’m not disputing what you said Danny. I’m disputing the way you went about trying to prove your point."Still don’t believe me, that PageRank isn’t the most important thing when it comes to ranking well on Google? Here’s a way I’ve been proving it for years."The SERPs don’t prove PageRank isn’t the most important thing. It does show PageRank isn’t the only factor at play, just like running similar "tests" on keywords in TITLE will show keywords in TITLE isn’t the only ranking factor (just because you have [seo] in your TITLE doesn’t mean you’ll rank #1 for [seo] - and yet we advice people to optimize their TITLE tags). Beyond pointing out the obvious (that PageRank isn’t the only ranking factor), what does that years-old "proof" prove?
@halfdeck, I think you’re going way too much into depth about why I use that example.Look, I get people who say "I’m PR5, I should beat a PR4." So you want to show them that no, it’s not just that the page with the most PR wins.For them, it’s PageRank beats all. I try to illustrate that it’s a factor. Yes, perhaps it is the most important factor in the overall mix, but that other factors can still cause a lower PR page to rank better. But that’s not the leval most people are thinking about.
"I think you’re going way too much into depth about why I use that example."Man, ok, but frankly the example annoys me to no end because I see people repeating it over and over on forums."Look, I get people who say..."Danny, yeah, I get that. I know where you’re coming from. What I’m saying is your "proof" doesn’t prove anything except PageRank isn’t the only ranking factor. You’re claiming it also proves PageRank isn’t the most important ranking factor - your example doesn’t prove that. That’s all I’m saying. Am I mincing words? Maybe. But SEOs have a habit of drawing the wrong conclusions based on a given set of facts (e.g. subdomains will be treated as subdirectories, second link never passes anchor text, knol pages receives an automatic ranking boost, googlebombing algo is bugged). I don’t see anything wrong with encouraging people to think with a little more care so they think twice or three times before jumping to conclusions just because they want to believe something is true.
Danny, halfdeck, seth, darkmatter, anyone...!Don’t folks think it strange that Google isn’t indexing the vast bulk of knol content? I wrote up what I’m seeing in a bit more detail: http://web-owls.com/2008/07/29/google-knol-the-new-invisible-web/I’d really like to hear your thoughts on this. Thanks.
dsarokin, your experiment’s results doesn’t surprise me at all. You’re doing closer to a random sample on knols, and it’s utterly believable to me that none that you’ve tested currently rank. Knol pages are regular web pages and they should (and I expect them to) rank depending on whether someone links to them as well as on-page factors like keywords, title, etc. Danny picked an interesting example in that he looked at all the knols on the front page, which received not only links flowing PageRank from the front page of knol.google.com, but also quite a few links from lots of people who were writing about knol.<div></div><div>If you look back at Danny’s "The day after" article, when Danny mentioned the searches he tried, he linked to the knol pages with that anchortext, not to the Google searches for those terms. So Danny’s write-up is partially contributing to the knols that he’s writing about continuing to rank highly. As our crawl/indexing system has adjusted so that we know knol.google.com shouldn’t get any of the authority of google.com, knol rankings will vary. For example, I see the mostly often-discussed (and linked to) knol about "How to Backpack" ranking #6 right now. But I expect most knols would be much closer to your experience--they aren’t getting a bunch of links from people talking about them, so of course they aren’t ranking highly.</div>
LMAO.. patience.Thanks Matt Cutts
By the way, I’ve been looking into this issue more tonight. I thought I’d take the "how to backpack" knol that was featured on Knol’s launch and dig into that case. That specific knol has over 100 links pointing to it. I see plenty of reputable sources, e.g. http://www.washingtonpost.com/wp-dyn/content/article/2008/07/23/AR2008072302098.html?nav=hcmodulehttp://www.alleyinsider.com/2008/7/google-launches-wikipedia-killer-knol-everything-you-wanted-to-know-about-toilet-clogs-tooth-pain-more-goog-I saw plenty of techie and news sites pointing to that specific knol, e.g.http://content.zdnet.com/2346-9595_22-212076-10.htmlhttp://technologizer.com/2008/07/23/knols-well-googles-encyclopedia-looks-cool/http://www.webpronews.com/expertarticles/2008/07/24/google-knol-launchesI saw regular bloggers linking to the backpack knol, e.g.http://www.greghughes.net/rant/KnolGooglesNewPlaceToExplandYourKnowledgeOfWellEverything.aspxI even saw a bunch of international sites linking to this knol, e.g.Russian: http://www.lenta.ru/articles/2008/07/24/knol/Polish: http://ireporter.blox.pl/2008/07/Knol-czyli-Google-dla-wolnej-kultury.htmlCroatian (I think!): http://tihomirs.blogspot.com/2008/07/knol-google-ova-online-enciklopedija.htmlRight now I see that knol ranking at #6 if you query for the title of the knol (how to backpack). It’s ranking there because of its own merit (on-page keywords as well as off-page backlinks), not because of the authority of google.com. When knol.google.com launched, a lot of people linked to the front-page knols as they were writing up their reactions, so that certainly helped that first set of knols. But my experience closely mirrors that of dsarokin. I don’t see the backpacking knol in the top 100 for the query [backpack] or [backpacking], for example.I repeated dsarokin’s experiment by looking around my place. I’ve got Rock Band drums, a quilt, and a bike within sight. I did a search on knol for [drums], [quilt] and [bike] and selected three knols: Title: Sonor drums, url: http://knol.google.com/k/rob-van-putten/sonor-drums/3j3t0x6beqpop/3#Title: T-shirt Quilts, url: http://knol.google.com/k/beth-sullivan/t-shirt-quilts/3lrmtn3usod1n/2#Title: Bike Across Nevada, url: http://knol.google.com/k/john-schwaner/bike-across-nevada/35rurl1jm04qb/4#I did the queries [sonor drums], [t-shirt quilts], and [bike across nevada], then I checked the top 100 search results. No knol urls were in anywhere in the top 100.Just to reiterate, if you create a knol you won’t get the trust/authority of the google.com domain.
thanks for the infos and the testing Matt Cutts.
Who cares if Google gives Knol a pagerank advantage? I certainly do not. Unless, Google allows participants to SPAM Knol, how does it benefit anybody on Sphinn?
Matt, I guess I’m not being clear. I’m not talking about knol articles not getting a good ranking in search results. I’m talking about them not being indexed. Period. They’re not simply low down in the search results. They don’t show up in search results at all because they don’t exist in Google’s index. The pages are never spidered (or at least, haven’t been in the week or so since Knol’s been public).Maybe I’m being unrealistic. But I’m used to new content (from anywhere...even the most obscure sites) showing up within a day. An exact-phrase search on a telltale phrase from new content will show up in search results very quickly.But none of the Knol content that I looked for (with exact phrases) is showing up. It’s just not being visited by the Google bot. I don’t know how else to put it, but I hope I’ve clarified a bit.
@dsarokinthey probably aren’t getting indexed because they don’t have any links, unlike the examples Matt showed. Usually, a page doesn’t get indexed until it’s discovered through a link on an indexed page.
"I’m talking about them not being indexed. Period. "Indexing issues are usually about lack of PageRank (though you can get a few pages indexed using RSS and zero backlinks). No backlinks = no gass = car stuck on the side of the road.
Right dsarokin. They’re probably not getting crawled right now because they don’t have any links.
Matt, as you saw from my article, I dug into the backlinks and couldn’t find much. And I couldn’t find much at that time because (1) it was a week ago, so there are new links that have emerged since then and now from when you’re looking and (2) more important, Google wasn’t reporting much. If Google itself would just fully report links, as I pointed out in the story, you wouldn’t need to debunk. I or others could do it ourselves.Here’s weirdness. That pack was ranking at 19 for "backpacking" a week ago with fewer links than it has today -- and today, it’s like not in the top 50, from what I can see. So it has all these new links now that it didn’t have the day after (or which weren’t reported properly by Google) and it’s ranking worse?The main thing seems to be that it’s not listed on the Knol home page itself, I’m guessing. And that goes back to what I wrote -- that the Knol home page has a lot of PageRank authority that in turn helps pages listed on it do better. OK, so those will come and go. But I keep coming back to the fact that Google has created this problem for itself. Google didn’t need to do Knol, and each time a Knol page ranks better, some folks are going to assume overt favoring. And other folks are going to understand there’s just natural ranking that happens for all the right reasons but it’s still Google that takes up another slot in the search results (and yep, I know it’s actually the author getting the spot, and they can get the AdSense revenue in part). But it still leads back to a Google property.
an influx of links, good rankings, then over time adjustment of link value and consequent decrease in locations... now what does that remind me of ?
"If Google itself would just fully report links, as I pointed out in the story, you wouldn’t need to debunk. I or others could do it ourselves."The lesson simply is don’t assume. Whenever we publish something we put our credibility on the line. That’s the gamble we take. But if you turn out to be wrong, don’t blame Google. You can avoid being debunked by leaving a back door open. A title like "knol does indeed give pages a ranking advantage" is difficult to back away from.Obviously people are afraid of knol, Wikipedia, and YouTube taking up 3 slots on every SERP. But the fact that Google is wrong to release knol doesn’t justify publishing false claims presented as fact backed up by blind fear and sloppy testing.If you don’t like knol, why not just publish a post showing a million ways how knol blows? All you’ve done with this post is create FUD.
My title also was: "The Day After: Looking At How Well Knol Pages Rank On Google." It’s not the title of what was here on Sphinn. My opening paragraph was, "I came away feeling that being on Knol does indeed give pages an advantage they might not get if they’d been hosted on some other brand new web site."You can disagree with that, but that was my feeling -- qualified as my feeling. Indeed, further qualified in the story:"There are lots and lots of caveats to consider. For one, doing a backlink lookup in Google is pretty useless...."My issue with Google undereporting links is long-standing, and there are many examples where because they fail to do this, people can make mistaken assumptions. I think Google should fix that, and I think they do share part of the blame if there are mistaken assumption out there because of this.
"It’s not the title of what was here on Sphinn."My bad Danny. I didn’t notice."I think they do share part of the blame if there are mistaken assumption out there because of this."I don’t mind Google sharing the blame as long as SEOs don’t pretend they’re blameless. Besides, does Google gain all that much from clueing in SEOs (apparently Matt thinks so, since he provided input in this thread, but I’m not sure about the rest of Google)? I also don’t believe its that difficult for people to hold off judgement till a hole’s dug deep enough, though I understand when it comes to reporting news, we all want to be first. Report first, research later, right? But then again that kind of atmosphere makes it hard to believe anything, or anyone.At the end of the day, I think we share the blame in jumping to conclusions time after time. It just makes us look like noobs.
Danny, Google has made a conscious choice to show full backlinks to the owner of a site, and a sample of backlinks to the general public. I understand that some people dislike that policy, but that’s a completely independent issue. I believe that we only push new backlink data when we update toolbar PageRanks, so relying on the "link:" operator on the main Google search just a day after a site has launched will almost always report no links. For example, the search [link:http://searchengineland.com/080724-140223.php] doesn’t return any backlinks either, even though of course many people have linked to your story in the last week or so.<div></div><div>Responding to "Google didn’t need to do Knol" I think Knol can fill a really helpful niche, because not everyone wants to go to the trouble of buying a domain name or setting up a blog when they just want to put a piece of information up on the web. My concern is that even if you add up the two issues of "Google didn’t need to do Knol" and "Google should report full backlinks," that does not equal "Google is favoring its own content."</div><div></div><div>I know that you didn’t write the title of this Sphinn Danny, but that was clearly the impression that a few people came away with from the story. I just checked back here because of this comment: http://www.webpronews.com/topnews/2008/08/01/cuil-interest-shows-hunger-for-search-competitor#comment-34647 in which an anonymous commenter claims that "google is well on it’s way to favoring it’s own content over an impartial list from the web."</div><div></div><div>So now I’m faced with the task of debunking this misconception around the web. Of course, I pointed out that Yahoo is actually ranking the backpack knol higher than Google right now (#3 on Yahoo vs. #7 on Google when I looked at it). It’s funny that DaveN is ranking in one day for [how to backpack], but no one claims that Google is boosting DaveN in our search results. :)</div><div></div><div>Likewise, it’s worth pointing out that for the query [how to backpack], the #1 search result on Yahoo is actually a hard-coded shortcut to a Yahoo property due to a partnership with a dictionary company. Did money change hands for that shortcut? We don’t know in general with Yahoo shortcuts--all we know is that some Yahoo shortcuts are paid for, because as they mention on their shortcut page: "Some of the content may come from partners who pay to be included in Yahoo! or have another financial relationship with Yahoo!."</div><div></div><div>I understand that Google is getting more questions about this topic because Knol just launched. And I understand that it was natural to ask whether Knol is getting some unfair advantage by inheriting the trust or authority of google.com, but it’s not.</div>
Matt, I have ranked a Knol page in the top 20 for a fairly competitive term with a single link to it. Are you claiming that I could do the same thing with a brand new site or a page on one of my own existing sites?No offense, but even if you guys didn’t INTEND Knol to get a rankings advantage is not the same as what happens in practice. I realize you’re all geniuses over there but you’ve screwed stuff up before.Also, I notice that you say "unfair" advantage. Does that imply that some advantage is fair?
It’s interesting how people always want to assume that Google favors certain types of sites. Before Knol people thought Google gave an unfair advantage to DMOZ pages, and to .edu pages, and to wikipedia pages, and on and on.I wonder why it’s so difficult for people to believe that Google ranks them all by the same criteria?
"Danny, aren’t the results of this experiment affected by the freshness effect? That is, as you know, fresh pages rank better for a time after they’re created."It’s completely pointless to look at any traditional SEO rules in terms of Knol. Because where do you draw the line? If you are going to look at the rules, then why not look at them all. And if you’re going to look at them all, then Knol is ranking exactly as well as it should rank. It’s a subdomain of a website with the highest trustrank on the Internet. Can we agree on that? Do you think Google gives Google.com the highest trustrank of any site in the index. Yes, they do. If a person’s argument is that it’s unfair that Knol is ranking well, that argument has to be grounded in something other then traditional SEO rules because they all dictate that knol.google should be ranking extremely well out of the gate. And yes, with more backlinks overtime it’s easier to claim that they’ve earned those rankings, but the reality is, based on the rules of the algo, (which is that if you add pages on an extremely trusted site, they rank immediately ie. Digg pages) that Knol is ranking exactly as it should based on the existing rules.
"I would think stuff like this leads to the demise of Google’s trustworthiness. If nobody trusts Google, they’ll start jumping off the ship. Google is pretty impervious at this point, but not so much so that they shouldn’t have SOME fear people will start going elsewhere."The average user doesn’t care that Knol is ranking well. They like it, like they like the fact that Youtube is ranking well. So there’s really no fear that the masses are going to stop using Google because Knol is ranking well. <div class="comment-info"> <!--Reply--> <div id="reply-48955" align="left" style="display: none"> Send comment HTML is disabled <table border="0" cellspacing="0" cellpadding="0" width="320" height="240" class="mceEditor" style="width: 320px; height: 240px"><tbody><tr><td align="center"> </td></tr><tr><td dir="ltr" class="mceToolbarBottom" height="1" align="center"><img src="<img src="<img src="<img src="<img src="<img src="<img src="</td></tr></tbody></table> </div> </div>
"I wonder why it’s so difficult for people to believe that Google ranks them all by the same criteria?"Probably because if you make a Digg page, it shoots up immediately, even with 1 backlink. That would tend to indicate that certain sites (like Digg) have a different ranking criteria. Especially when the Digg page temporarily outranks the original artcile. I think it’s fairly obvious that sites have different ranking criteria Jill. Why is it that Techcrunch can push itself up the SERPS crazily by creating a few internal links... You don’t really believe that all sites have equal ranking criteria, do you?
Skitzzo, no--I don’t think there’s such as thing as a "fair advantage." We try to rank all our content on a level playing field. I’d be happy to look at your knol example if you want to share specifics. And Jeeb90, I’ve asked the universal search folks about their ranking and as far as I can tell, they bend over backwards to try to rank and return non-YouTube content. There’s just a lot of video content out there that is on YouTube, you know?<div></div><div>Jill, you forgot to include ".gov" links in the list <-- that was mild sarcasm :)</div>
Matt, I’ll go one more round to save the back-and-forth and give you the last say, if you want it....As for what Yahoo does -- this isn’t about Yahoo. Pointing out that for a query at Yahoo, they have a hard-coded shortcut doesn’t dismiss concerns people have about Google. It just distracts from the bigger issue as if to say, hey, if you think we’re bad, these guys are worse. I know there are plenty of people worse than Google. That doesn’t mean I want you worse at all (nor am I saying you are).Google’s decision to deliberately unreport links is not an independent issue. Because you do this, it is impossible for someone to tell from Google itself whether a page has much link authority pointing at it or not. I couldn’t do this for some of the knol pages I looked at, for reasons that I explained.It’s news to me that you only update link: data when toolbar PageRank data is updated. Since that happens on a roughly quarterly basis, wow. But then then that data is already of little use, so I guess it shouldn’t make much of a difference. But then again, it sounds like backlink data is updated more than a quarterly basis, because you were demonstrating how that knol page eventually did show more backlinks within a week.I think all this underscores what a problem Google’s underreporting of backlinks is, even for those who might want to help Google in debunking things about it. I wish you’d provide better reporting. That wish isn’t new. Yahoo manages to do it. Google could too. And I’ll leave it at that.In terms of the issues, I’m not saying Google is favoring Knol. I don’t believe I ever wrote that. People, as you know, will come away with whatever they want. I did have the impression that Knol was being seen as an authority site according to Google’s own ranking algorithms -- just like I feel Google sees many non-Google sites as authority sites, and pages within those sites seem to have in general a better chance of ranking than those on unknown domains.That’s not Google favoring itself. That’s Google favoring authority sites in general. But Google now seems to have gained another one, and as those pages rank, they inevitably may push other pages aside. And Google then looks to be helping itself, however it happens.And to further clarify, I’m not saying (and never did) that knol.google.com does well because it’s part of google.com. I’m saying that knol.google.com seemed to have, when I wrote this, quickly gained enough authority ON ITS OWN that pages within it did better -- that a page I never mentioned, which seemed to have practically no links pointing it -- shot to 28 out of 755,000 pages. Sorry, that’s just not something I think you’d see happen on most brand new sites. And again, not because Google did anything to favor itself. Just because the Knol site rapidly gained authority.In terms of the debunking you have to do, with respect, that blame lies not with my article but to the fact that Google ran this program at all. I’m not the only one who raised questions with ranking. And plenty of people will just assume regardless of articles that Google is favoring itself, if they see knol pages there. Heck, plenty already assume you must have a deal with Wikipedia, and you don’t even own them.The debunking you have to do is primarily due to the fact that Google itself has decided to run and operate a site that can have content on any topic and which inevitably raises these questions. It’s the same issue Google has to deal with in terms of AdSense -- that people will think Google favors AdSense sites. Google knowingly made decisions to do things that raise issues of inherent conflict. In terms of AdSense, I think Google is seen as having a good track record. In terms of Knol, you may well be found to have the same way.I’ll leave with these questions, which is what Jeeb90 stressed about Digg. It, like many sites, seems to have authority to its entire domain that helps any page within that domain do better than if they were outside of it and on a less trusted site. They’re not guaranteed to rank, but they have a better initial shot.Are some domains seen as "trusted" by Google, so that any page within them gains some of that trust in ranking mechanisms.If so, is this trust transmitted to subdomains of a domain?I think the answer to the first question is yes and the answer to the second is no. But I’d like Google to give us an on-the-record answer. It would help with the debunking.
Matt said: "Jill, you forgot to include ".gov" links in the list Jill, you forgot to include ".gov" links in the list"Yes, I did think of that while writing it, and that would fall into the same category.Danny said: "Are some domains seen as "trusted" by Google, so that any page within them gains some of that trust in ranking mechanisms."I’m certainly not Matt, but yes, of course some domains are seen as trusted. But not because of who they are or what they are but because they’ve earned that trust. So of course, pages on the google domain as a subdomain, i.e., knol.google.com will have whatever trust a subdomain on any other site would have. Only they have the power of their sub-domain being Google.But it’s not any more of an advantage than any other site that did the same thing.Does seem like this could backfire though, in that knol sites can be created by anyone and therefore not truly have the same trust factor that google.com has. I personally haven’t even looked at knol yet as I’m waiting for the hype to die down, but it sounds like from the other posts here that they’re using nofollow in all the links? If so, then that should lessen the problems that might be associated with anyone being able to create a page there.
"That’s not Google favoring itself. That’s Google favoring authority sites in general.""Authority" sites like Wikipedia have only one advantage: higher concentration of PageRank. Trust is just a way of monitoring PageRank flow. High Trust means juice flows undevalued. I’m not saying PageRank makes pages rank higher. I am saying higher concentration of PageRank means deeper index penetration. More pages in the main index = more power to internal anchor text and more Google real estate = insanely high rankings for thousands of terms.If knol.google.com’s link juice flows out to the individual knol pages so that the backpage page turns into a TBPR 6 page (like it often happens on Wikipedia) you got something to worry about. But otherwise knol page is just like any other page on a relatively frequently crawled/indexed site.BTW, I do believe "should Google reveal more accurate link data" and "is Google out of its mind in releasing knol" are off-topic to the question "do knol pages receive a auto-ranking advantage" to which I think now we all know the answer.
You folks are having a very interesting discussion. But from the point of view of someone providing content to Knol, I’m still really stumped. If I post a new page on eHow (where I contribute regularly), or on my own website or blog, the content is indexed and shows up on Google searches within hours.My Knol content, on the other hand -- with backlinks, ratings, and internal Knol cross-references -- remains totally invisible, even after a week. And again, I’m not talking about it being on page 5 or 10 of search results. I mean it just doesn’t exist in the Google index at all.It seems to me that, far from favoring Knol content, Google is doing just the reverse...keeping most of it out of view, and allowing only a few articles to appear in regular search results. Am I missing something?
dsarokin: I’d ask Matt Cutts if Google is using ’quality raters’ to review some/all(?) published knols: maybe one of them didn’t like your content?
"Am I missing something?"Yeah...links.Read the answers to your previous comment dsarokin; your question’s already been answered.
halfdeck.I’ve been reading all the comments...believe me.Here’s the thing, though. (1) New content at other sites I frequent shows up very quickly in search results, even without links. Links are critical for high rankings in search results. But I’m not sure links are as critical as folks seem to think they are to simply being anywhere in search results in the first place. That said, my knols have links. I’ve added them in several places: my web pages, some blogs, eHow, other knols, and right here: http://knol.google.com/k/david-sarokin/online-newspaper-archives/l9cm7v116zcn/5#This knol (about newspaper archives) contains the phrase: "archives generally take three forms". A search on that exact phrase should bring up the knol right at the top of the search results. But it’s just not there. Had I published that exact same article at any one of a number of other places, it would show up within hours.Kapish?
Halfdeck said: If knol.google.com’s link juice flows out to the individual knol pages so that the backpage page turns into a TBPR 6 page (like it often happens on Wikipedia) you got something to worry about.But that would only happen if the individual knol pages were linked to from knol.google.com. Since this doesn’t just happen without actual links to provide the juice, there shouldn’t be anything to worry about. Although, I do think that "trust" goes beyond just PageRank, and therefore, just being on the google.com domain, they have that one piece of the algo that other sites may not have straight out of the box.
"Had I published that exact same article at any one of a number of other places, it would show up within hours."Ever heard of internal links? There are plenty of places other than this thread where you can ask basic SEO questions like why a page isn’t getting indexed. It’s been answered three times here, once by Matt Cutts of all people. Take it or leave it."But that would only happen if the individual knol pages were linked to from knol.google.com. Since this doesn’t just happen without actual links to provide the juice, there shouldn’t be anything to worry about."Yeah, exactly."Although, I do think that "trust" goes beyond just PageRank"I oversimplied to make my point, and no doubt its more complicated than just PageRank, but the long and short of it is Trust is often about link manipulation and (internal) PageRank is Google’s standard tool for addressing that problem. I personally don’t believe a page will rank any higher just because its on a "trusted" site.
Halfdeck,Thanks for the direct reply. But Matt was mostly focused on the topic of results rankings, not the complete absence of indexing. To wit:"...I repeated dsarokin’s experiment by looking around my place. I’ve got Rock Band drums, a quilt, and a bike within sight. I did a search on knol for [drums], [quilt] and [bike] and selected three knols... I did the queries [sonor drums], [t-shirt quilts], and [bike across nevada], then I checked the top 100 search results. No knol urls were in anywhere in the top 100."I have no expectation of my knol making the Top 100. But I do expect it to at least exist in the Google index. On this matter, Matt said:"Right dsarokin. They’re probably not getting crawled right now because they don’t have any links." But they do have links...quite a few at this point. And I repeat, much content gets indexed quickly even without links. I know this for a fact by the speed that some of my own content gets crawled, with no internal or external links at all. Links are essential for ranking, but not -- as far as I know -- for being crawled by the Googlebot.I really think this whole thread should be retitled as: Knol Does Indeed Give Pages A Google Ranking Disadvantage.
"But they do have links...quite a few at this point. "Not all links are created equal. Just because you got 1,000 links pointing at a page means shit if Google doesn’t think much of those links."I know this for a fact by the speed that some of my own content gets crawled, with no internal or external links at all. Links are essential for ranking, but not -- as far as I know -- for being crawled by the Googlebot."Though there are several exceptions, in general, PageRank is the primary factor Google uses to control crawling behavior. That means links. If that’s news to you, you got a lot to learn. If you don’t believe that, your loss. Come back to this thread in 12 months and see what you believe then.
"Who cares if Google gives Knol a pagerank advantage? I certainly do not. Unless, Google allows participants to SPAM Knol, how does it benefit anybody on Sphinn?"You should care. I have to say I have not yet seen any spammy knol rank, but definitely there is a nice advantage in it. I also do see a nice structure advantage in how those knols are set up which may benefit in addition, but overall the domain seems to be what does it since I have yet to see anything similar when it comes to a domain name rank above others for no appearant other reason.It’s very interesting seeing this and getting an idea where the social communities will shift in the future.If Google does a great editing job and knows how to keep spam out, more power to the knol.If it becomes another blogspot mess, G__ help us all! :)Mike Dammann
halfdeck, or anyone,Here’s what I’m talking about. I made a Clipmark of one of my knols:http://clipmarks.com/clipmark/79395E4F-8DCD-4182-9F5A-907634BE75CB/There are no links to this article. It took about an hour to show up in Google search results:http://www.google.com/search?num=100&complete=1&hl=en&q=knol+%22newspaper+archives&aq=fWhy aren’t knols appearing with similar ease? If Google automatically crawls each new Clipmark, why isn’t it doing the same with each new knol?david
"There are no links to this article."There are links to the article if you look hard enough.
Yes. And there are links to all my knols. So why is my Clipmark spidered within an hour, and my knols are absent, even after a week?You seem to have a lot of experience in these things. I’d really appreciate the benefit of your insight.
It’s late enough that I’m not going to do a full reply to Danny or dig into dsarokin’s situation to analyze that. I did want to stop by to correct something that I said earlier: "I believe that we only push new backlink data when we update toolbar PageRanks, so relying on the "link:" operator on the main Google search just a day after a site has launched will almost always report no links."<div></div><div>I talked to the folks that push this data, and it turns out that we push this data closer to monthly. We might update data for the "link:" operator more frequently in the future, but not at the frequency where you’d expect to create a new page one day and be able to do backlink queries on it the next day.</div>
Matt Cutts, if you are still here :) could you kindly indicate if you guys use any ’editorial forces’/some kind of ’human quality rating’ for knols? Really appreciate a word :)
"And there are links to all my knols."Didn’t I say a few comments back not all links are created equal? There’s a PageRank threshold to getting a page indexed. It doesn’t matter if you got 1000 backlinks. If they don’t pass enough juice collectively, there’s a good chance the target URL isn’t going to get indexed.What links do you know are pointing at http://knol.google.com/k/david-sarokin/online-newspaper-archives/l9cm7v116zcn/5#, for example? All I see are two nofollowed links from Sphinn and eHow.What appears to be a link on this page isn’t an actual link:http://clipmarks.com/clipmark/79395E4F-8DCD-4182-9F5A-907634BE75CB/Also the Clipmark site has an RSS feed - you can sometimes get orphaned pages indexed using RSS.
dsarokin, yes, you’ve got a strange situation. I did this search:"online newspaper archives" site:knol.google.comIt didn’t bring up your online newspaper archives page at Knol. Odd, since it really should show for that search. Yahoo doesn’t bring it up either.My guess is that Google hasn’t found any non-nofollow link to it, so hasn’t spidered it or spidered it in a way that says it should be indexed (I get confused these days -- if Google only sees a nofollow link to a page, whether it will still index that page anyway if the page itself has no blocking. Maybe).
jmaulson, right now we have people checking Knols for spam (just the same as we look into reports of spam on other pages), but otherwise nothing like editorial or ratings.<div><div></div><div>Danny, if we see nothing other than a nofollow link to a page, we won’t index that page. Going back to your longer comment, a few thoughts:</div><div></div><div></div><div>- Faster or more comprehensive updates of link: data would help you in debunking some times, but it would also be abused by people doing things like trying to scrape all the links of a competitor. I think the Microsoft experience made that point where they had to disable link:-ish queries completely because of heavy scraping and bot activity. And even if you had perfect backlink data from the link: command, unless the rest of Google’s scoring was 100% open, there would always be cases that you couldn’t fully debug or dig into.</div><div></div><div></div><div>- Regarding the fact that a knol showed up in short order on page one for e.g. "how to backpack," it’s worth noting that DaveN’s post about the subject also leapt to page one for "how to backpack", also in one day. If you do the query as a phrase quote, there’s only ~35K results for that phrase, so it’s not a very competitive phrase.</div><div></div><div></div><div>- You mention that part of the issue is that Google has launched Knol, and so we’ll naturally be subject to questions about whether Google is somehow favoring Knols in the same way that people ask that question about AdSense. That’s a fair point, but just because people will ask that question doesn’t mean that we shouldn’t have launched (say) AdSense. Does it raise the bar in terms of messaging that we need to do to convince people that we are being fair in our rankings? Yup, I would say so. But is it still worth launching products like AdSense and Knol? I would say yes. Take AdSense for example. It really has enabled a whole wave of people to begin publishing their own site and monetizing that. Do some spammers use AdSense? You asked Eric that question, and his answer was yes, but that spammers in AdSense are a success problem: AdSense works well enough for most people that a few bad actors try to abuse it. But that doesn’t mean that AdSense wasn’t a good thing to launch or that it wasn’t an over net positive for small publishers and the web. (And by the way, it’s gotten a lot tougher for spammers to subsist on AdSense as we’ve gotten better at finding them, throwing them out of Google’s index, and throwing them out of AdSense completely.)</div><div></div><div></div><div>I think it’s absolutely fair and right that you and others raise these questions. I think our crawling/indexing/ranking handles knols well, e.g. doesn’t give them any extra trust or authority just because they happen to site on google.com. And I do think that Knol sits in an interesting area on the spectrum of content creation. They’re easier to create than many blogs, because blogs can keep sucking your time away forever. :) But knols are not just throwaway pages either, because an author has some ways to prove their identity and establish their reputation.</div><div></div><div></div><div>So Knol may succeed or fail, but I’m glad that we’re giving this a try just like I’m glad that Yahoo is trying Yahoo Answers. I think Knol would be great for how-to type pages, for example. Or for memex-like summaries of "what I learned about subject X after researching it, because I didn’t find a nice capsule summary anywhere else."</div><div></div><div></div><div>In fact, I just did a blog post about JavaScript bookmarklets on my blog. I was looking around for information on them and wasn’t happy with what I found, so I collected 2-3 hours of digging and summarized what I learned into one blog post. If I’d thought about it, I would probably have put the information into Knol instead. I’ll have to keep trying to make the mental switch. :) But regardless, for someone who didn’t have a blog but was doing JavaScript bookmarklet research, a knol would be the perfect place to write up what they learned. The net result is that some new content is created. That content is open to everyone and makes the net more useful for the next person with a similar interest.</div></div>
Thank you much for the answer, very good post here, I may start reading your blog Mr Cutts :)
Knol is essentially Google’s attempt to SEO existing content and call it their own, sure it’s a good thing we can all get involved and make some money.There things I could do to your search engine in an afternoon that would generally add value, republishing old content isn’t one of them.What about the 3 trillion pages of content sat on Google’s servers, we’ve recently become aware of through Google’s own spokesmen, who has access to them? As a publisher and SEO, what about the 300+ pages of original content cached by Google but not indexed. What’s has happened to that knowledge? Who now owns that? Is it being distributed. No?The web is like a compost heap; it’s all the little beasties that keep it alive, putting a giant sheet of plastic on top like Google are doing with manual alteration of SERP’s, holding back genuine content, misinforming webmasters, publishing using duplicate content, whilst watching genuine publishers go to the wall, stinks.You stand for all you stand against, whilst hiding behind the viel of something you’re not.Your time will come Google.
Update (for anyone interested).My knols are finally showingup in search results, at the very same time that Google ’cleared’ my pages for Adsense ads (there was some sort of 2-week verification process). I have to believe that the two events are directly connected, though what the connection is, is hard to say. Thanks to all for listening to me blather, though.David
Follow-up blog post:Google Knol Rankings Revisited
dsarokin, please don’t jump to that conclusion. AdSense is completely independent of how/whether we crawl a particular page and does not affect our indexing or scoring in any way (positively or negatively). If we crawled your knols, it was most likely because we saw some links to them, but it would have nothing to do with AdSense or getting approved for AdSense. :)