| 8:13 pm on Feb 12, 2011 (gmt 0)|
The article is worth reading - through all five pages there are some good tidbits. For example, about the recent algo update:
|Google had just begun to roll out an algorithm change that had a negative effect on Penney's search results. (The tweak affected "how we trust links," Mr. Cutts said, declining to elaborate.) |
On Feb. 1, the average Penney position for 59 search terms was 1.3. On Feb. 8, when the algorithm was changing, it was 4. By Feb. 10, it was 52.
| 8:20 pm on Feb 12, 2011 (gmt 0)|
I get a login screen for that article. Too bad.
| 8:36 pm on Feb 12, 2011 (gmt 0)|
ken_b - the article is loading for me. However, if you are getting a paywall, just type "The Dirty Little Secrets of Search" into Google and then click through from the search engine and you are in (paywalls don't apply to people who can Google!).
On topic - I guess the comment quoted by tedster above confirms that the Feb 8th algo change was about links (the change that the Google employee posting on that Hacker news thread aluded to (and quoted in this webmasterworld thread:
I'm guessing the other changes mentioned in the following thread haven't been released yet?
(edit: just realised I put the wrong link for the second thread, now corrected)
[edited by: AlyssaS at 9:22 pm (utc) on Feb 12, 2011]
| 8:46 pm on Feb 12, 2011 (gmt 0)|
I thought it was interesting that the algorithmic change is described as taking so long to roll out. We have been noticing for months how big algo changes first show up and then seem to bounce around for days. This article seems to point in the same direction. I'm thinking statistical sampling and feedback - machine learning of some kind.
It seems that "how we count backlinks" has been a common description of algo changes for quite a while now. Makes me wonder if Google's address to content farms will also be in this area. And also if a site is tagged as hosting a low value outbound link, might that roll over to other outbound links they host.
I know that Google can negate link equity flow at the level of individual link, the whole page or the whole domain. But I wonder if enough "individual problem links" could spell bigger trouble for the hosting site.
[edited by: tedster at 8:57 pm (utc) on Feb 12, 2011]
| 8:55 pm on Feb 12, 2011 (gmt 0)|
Well, when they announced the scraper update in Jan, they mentioned the following:
|The new classifier is better at detecting spam on individual web pages, e.g., repeated spammy words—the sort of phrases you tend to see in junky, automated, self-promoting blog comments |
So - if they detect any of those spammy words on a page, they may assume that all the other links on that page are spam too. Perhaps they also have a filter based on how many links there are on the page too.
It lines up with what the NYT article said about that blog owner who got interviewed - I think he had an entire page full of links he was being paid for, and perhaps one of those links had the spammy words and doomed the whole lot.
| 9:42 pm on Feb 12, 2011 (gmt 0)|
A couple of thoughts come to mind on this thread.
First one - NYT doesnt know the half of what goes on - calling that "dirty little secrets" clearly shows how far behind the rest of the world they are. If NYT only knew the half of it :)
Second - yes - it would seem as though its a link trust thing of quite some complexity - or at least fairly hefty processing they are rolling out.
Which gives rise to the third - if they can go to all that trouble to get shut of so termed "spammy links" - then how come they cant do this :-
' If the link is on a forum profile then dont pass juice.
If instr$(URL, "profile") > 0 and instr$(URL, "forum") > 0 then
Its a spam link
The mind boggles - theres such a large volume of spam links out there ready to be zapped with one quick test its incredible that they dont focus on it for 10 mins or so.
Well - you get the picture - they'd do it through the inbound link profile probably but not really that hard.
| 9:50 pm on Feb 12, 2011 (gmt 0)|
As a forum administrator, I'm certain that forum profile links are far from all spam. There's important data in forum profiles around the web that Google wold be foolish to ignore.
I think the link weight issue is much more complex than even we can guess - and if we saw all the data that Google engineers have, we would all be boggled. The scale is immense, as is the data churn.
| 10:03 pm on Feb 12, 2011 (gmt 0)|
|Google and then click through from the search engine and you are in (paywalls don't apply to people who can Google! |
Thanks! But apparently they do apply when you try to go to the second page.
Oh well, I imagine it'll get discussed enough that the crucial points will get out.
| 10:18 pm on Feb 12, 2011 (gmt 0)|
forum links are important! Many popular products, sites and services are discussed on forums. If something has any value, it will likely be talked about on some discussion board. Its more about discrediting spammy links on forums which I am sure google already does. Not to mention, legit forums moderate links well and any non useful/spammy links will be removed.
| 10:23 pm on Feb 12, 2011 (gmt 0)|
I'm having no problems accessing the whole article.
Here's a quote I love, from the second page:
|When you read the enormous list of sites with Penney links, the landscape of the Internet acquires a whole new topography. It starts to seem like a city with a few familiar, well-kept buildings, surrounded by millions of hovels kept upright for no purpose other than the ads that are painted on their walls. |
| 10:43 pm on Feb 12, 2011 (gmt 0)|
This bit is interesting too:
|On Feb. 1, the average Penney position for 59 search terms was 1.3. |
On Feb. 8, when the algorithm was changing, it was 4.
By Feb. 10, it was 52.
The Feb 10 bit is the manual penalty. The algo change only moved them down a little bit, which indicates that G is being cautious.
Perhaps what's happened is something like this: they detect some spammy words on a page, are not sure whether that means there is an absentee blog owner/lack of moderation, or if the links are paid, and decide to only assign half the weight they normally do to that page, so the "juice" that is passed on is diminished.
If all your links are from pages like that, you suffer badly, but if you have some other decent links, you still rank, just not as well as before.
| 10:52 pm on Feb 12, 2011 (gmt 0)|
Company a does not like company b, so company a digs some dirt up and sets up an interview with a single google employee who makes the decision to destroy and ruin company b's rankings. Is that not evil or what? That right there shows that a few people in google have a lot of power that could be abused.....
| 11:12 pm on Feb 12, 2011 (gmt 0)|
|"...and they found a large number of backlinks, strewn around thousands of nearly abandoned websites." |
"Company a does not like company b"
Here's a paranoid interpretation: Company A finds abandoned websites, sets up the worthless backlinks, gives Company B a nice boost, then Com-A reports Com-B to Google as abusing their Terms, and sits back with a smile as they drop like a stone.
Oh that's right, there's almost nothing that another company can do to hurt someone's ranking. "Almost"...
| 11:36 pm on Feb 12, 2011 (gmt 0)|
|Company a does not like company b, so company a digs some dirt up and sets up an interview with a single google employee who makes the decision to destroy and ruin company b's rankings. Is that not evil or what? |
It's not a lot different to a webmaster-joe submitting a spam report through Google really though and that's a service we (mostly) all appreciate having.
That said, if the US press are anything like the UK press I treat most of their "exposes" as self-serving sensationalism with only one purpose and no regard for who it hurts.
| 11:38 pm on Feb 12, 2011 (gmt 0)|
|Here's a paranoid interpretation: Company A finds abandoned websites, sets up the worthless backlinks, gives Company B a nice boost, then Com-A reports Com-B to Google as abusing their Terms, and sits back with a smile as they drop like a stone. |
Yeah, but if they simply discount the links then it's not a big deal and a complete waste of time and energy on the part of Com-A, which seems very fitting in a case where someone else intentionally tries to damage the rankings of a competitor...
| 11:48 pm on Feb 12, 2011 (gmt 0)|
I found this interesting....
|Google had detected previous guidelines violations related to JCPenney.com on three occasions, most recently last November. Each time, steps were taken that reduced Penney’s search results |
Especially in light of this...
|A Penney's spokesperson denied that they had authorized those links - and they fired their search consulting firm. |
Apparent JCP and Google had crossed paths before. I wonder what those "steps" that were taken earlier by Google were. I also wonder if the firm who placed these links for JCP was the same one that was involved in the previous efforts.
[OK, I had to turn on cookies and accept 18 from the NYT site before I could get past page 1.]
| 12:01 am on Feb 13, 2011 (gmt 0)|
Well, I guess JC Penney will now have to spend that money on AdWords to get to the top of the search results...
We would not want them to spend that money on links to get to the top of the serps, that's cheating...
| 12:13 am on Feb 13, 2011 (gmt 0)|
If these types of links are simply discounted or ignored, then they have nothing to worry about, since in theory they should have a considerable number of great, quality editorial links in place that Google would love :)
| 12:20 am on Feb 13, 2011 (gmt 0)|
I just got a hold of an exploited site where they were doing SEO with hundreds of files in a folder (not JCPenney). This seems to be a big business. It's unfortunate for people who are doing things honestly.
| 12:28 am on Feb 13, 2011 (gmt 0)|
dvduval - out of interest, did that folder have spammy words in it? In other words, if you were a machine and not a person would there be words there that would indicate it's not just a resource page with lots of links?
| 12:49 am on Feb 13, 2011 (gmt 0)|
Google should have discounted the links and not made any public statements what so ever. What is so special about the new York times anyways?
| 1:21 am on Feb 13, 2011 (gmt 0)|
The sad part about this is, it is being played out in public.
Will JC Penny get a long multi-year penalty - no.
Will JC Penny have to apply through the web master console for inclusion - no
All they have to do is "fire their SEO firm" - wink wink.
| 1:52 am on Feb 13, 2011 (gmt 0)|
If it took all that to level JC Penny you can imagine the thousands escaping the noose. Plus how many "average joe's" would have access to MC. I would say the vast majority of the "big boys" are pretty safe no matter what they do. Google was on the line publicity-wise and legally if they did nothing. Do you think you would only drop to the 50's if you did the same? How long do you think JC Penny will be in a time-out? It's Wall Street versus the little guy all over again. We apologized and supposedly fired someone. Now put us back in our locked-in positions.
| 2:04 am on Feb 13, 2011 (gmt 0)|
What surprises me about this is how easily Google was hoodwinked. With all their resources, and all the time they've been working on search, their algorithms should have been able to detect an artificial linking pattern of this magnitude.
| 2:49 am on Feb 13, 2011 (gmt 0)|
Yeah - that's essentially what Matt Cutts said, too. I'm also amazed that no flags were raised on this.
About that fall to #52 - notice that was the average position for 59 different query terms that previously averaged #1.3. So it's not losing just one ranking. Even though we don't have the list, my guess is that nothing like that level will be achieved again so widely.
Sure, they'll achieve some decent rankings again, just like BMW did. But here Penney's were stealing the game in a very big way. It's very unlikely to see them again at #1 for [dresses] and [bedding] and [arearugs] and [Samsonite carry-on luggage] and 55 others.
| 2:55 am on Feb 13, 2011 (gmt 0)|
Since Google apparently had flagged them 3 times before I would have thought that they would have had a "watch" on JCP.
| 3:32 am on Feb 13, 2011 (gmt 0)|
|But here Penney's were stealing the game in a very big way. It's very unlikely to see them again at #1 for [dresses] and [bedding] and [arearugs] and [Samsonite carry-on luggage] and 55 others. |
And they care? Take the money and run... They weren't there before (afaik)... They were there for the holiday season... They're not there now... What do they care? One holiday season at number 1 for those 59 terms when they probably weren't even close before (and aren't again) isn't too bad a move, imo.
How much do you think they made over the holidays from ranking at the top for those terms? Really, what do they care now that the $ days are over and they're probably close to where they were before they gamed the system?
If someone told you they could plant you at the top of those terms for the Christmas shopping season but it might not last, would you go with them? 1 or 2 = $ :: 11-1,000,000 makes about no difference. So they're 52 ... Traffic isn't that much less than 12 is it? Who even goes to page 2?
| 3:56 am on Feb 13, 2011 (gmt 0)|
Google's actually WAY behind the game in some ways IMO... They're determined to do everything algorithmically, not manually, so what happens? People game the system for the money terms during the money time and take the cash to the bank while the people at Google try to fix it so it won't happen again, but of course while they're fixing one 'game' someone else is gaming them in a new way and once it's discovered it takes months to fix, again.
Who's ahead and who's behind in the game of SE rankings?
| 5:28 am on Feb 13, 2011 (gmt 0)|
I mentioned a month or two ago another big damaging google article was in the works.
Btw wait until you see the next one about unread spam reports and duplicate sites being set up to purposely embarrass google.
The other big game is setting up shadow sites that run blatant duplicate content using different keywords as I have mentioned. Apparently it really does work and even if it's reported by bundles of users nothing gets done ;)
Google should be applauded for what they did for all of us and the internet but they deserve all the heat they are getting as they were asleep this last year or two.
How hard is it to detect a duplicate site when it uses the same page naming structure, same chat tags and is owned by the same company? They can trap my wifi network and assign it to my house when a proud drives by but can't see that xyz Corp is selling the same item on ten sites with different keywords?
You know what is most infuriating? All the legitimate businesses hurt by jc Penney just got helped by the times. We all get nothing because we didn't get a piece written on our segments. May have saved my job.
| This 158 message thread spans 6 pages: 158 (  2 3 4 5 6 ) > > |