homepage Welcome to WebmasterWorld Guest from 54.161.190.9
register, free tools, login, search, pro membership, help, library, announcements, recent posts, open posts,
Pubcon Platinum Sponsor 2014
Home / Forums Index / Google / Google SEO News and Discussion
Forum Library, Charter, Moderators: Robert Charlton & aakk9999 & brotherhood of lan & goodroi

Google SEO News and Discussion Forum

This 138 message thread spans 5 pages: < < 138 ( 1 2 [3] 4 5 > >     
Signs of Fundamental Change at Google Search
tedster




msg:3302704
 12:07 am on Apr 5, 2007 (gmt 0)

In the 950 penalty thread [webmasterworld.com], randle recently posted that "...there is a broader picture here, fundamental changes are afoot."

I agree, and I'd like to collect people's observations about what "signs of fundamental change" they've may have observed in recent months.

One big one for me is that no matter how common the search term -- it could generate billions of results -- you always seem to bang into an "omitted results" link before your reach #1,000. In fact, i just checked out a search on the word "the" which google says generates 5,300,000,000 results. And even on this monster, #928 is the "omitted results" link. Hmmmm....

Now 5,300,000,000 also seems like a low number to me - unless it does not include any Supplemental Results. So my current assumption is that by fattening up the Supplemental Index, Google has pared down the main index to somewhere in the vicinity of 5-6 billion urls.

A related sign of fundamental change, I feel, is the problems Google currently has generating understandable results for the site: operator or the Webmaster Tools reports. It looks to me like the total web data they've collected is now broken up into far-flung areas of their huge server farm -- making it very difficult to pull comprehensive site-wide information together again.

 

annej




msg:3304673
 12:23 am on Apr 7, 2007 (gmt 0)

Steve, Are you saying we should not vary our anchor text now? I like using the same terms in the navigation throughout the site as otherwise the visitor won't know that I am refering to the same page. Is this a no no now?

steveb




msg:3304802
 6:35 am on Apr 7, 2007 (gmt 0)

The comment mentioned the danger of varying anchor text.

Many WebmasterWorld members have talked about varying anchor text as an optimization tactic for a couple years, so it would be no surprise if Google (tried to) attack that.

Regardless of Google's response, the best thing to do with anchor text is write it naturally for users, linking to pages with the words of what the page is about. This means normally consistent anchor text in navigation and everywhere, with a little bit of variation when linked to contextually.

That would seem like it should be safest with Google, but as we see, Google can screw anything up, so nothing is safe on the one hand and nothing is consistently "most effective" either.

Petra Kaiser




msg:3304882
 10:54 am on Apr 7, 2007 (gmt 0)

... signs of fundamental change ...

I want to add our following observation to this interesting discussion.
For our pages, search on higher competitive phrases in the +1.500.000 results range, is affected (down from page 1 to page 9). Less competitive phrases, search in the -500.000 results range, is not affected, SERPS still remain constant for all pages.

tedster




msg:3305016
 3:47 pm on Apr 7, 2007 (gmt 0)

The paraphrased comment from the lady at Google mentioned navigation: "remove multiple keyword variations from nav menu links" - but that is not the same thing as talking about inbound link anchor text from other sites. When backlinks are natural, then variation in anchor text would also be natural and expected - and most likely unavoidable.

Still, I can see how the comment is open to various interpretations as written. Does it mean don't change the nav menu's anchor text when it appears on different pages? ...or does it mean that even in one page's version of the nav menu, we should be wary of repeating the same keyword in multiple phrases?

I assumed the second, but it certainly can be read as meaning the first. I definitely would not reccomend that - not just for Google, but also for users. Consistency in labeling is an important part of helping visitors stay oreinted as they browse a site.

annej




msg:3305113
 7:10 pm on Apr 7, 2007 (gmt 0)

When you think about what is natural it is natural to have the anchor text in your navigation menu the same throughout the site. It is NOT natural that everyone who links to you would use the same anchor text.

It seems to me they are two different situations and one would hope that Google sees them differently.

rekitty




msg:3305154
 8:20 pm on Apr 7, 2007 (gmt 0)

Great thread. Yes I believe there has been a fundamental change in Google. Tedster, I believe your re-ranking is a major change, but not the fundamental change.

The fundamental change I see is Google has significantly altered the calculation of link popularity by relegating a large percentage of pages to the supplemental index where their links don't count.

Previously every page on the web had a vote as long as that page was found by googlebot. If a page was not found for a couple months or if it had duplicate content it was put in the supplemental index. Supplemental pages didn't get votes because they could not be trusted. Not a big deal because only a small percentage of the web was supplemental. The supplemental index was primarily the garbage collector where pages went to wither.

Today is very different. The supplemental index plays a key role in Google's new architecture and a huge percentage of pages are sent to the supplemental index for many different reasons. These pages no longer rank well, making some noticeable changes in the serps. The huge effect, though, is all these supplemental pages no longer pass link popularity from what I can see.

I think of it as an election where half the voter registration cards have been torn up. The election results will be very different. Did only folks over 40 get to vote? Or was it a particular income level or religion that voted? Who you let vote will result in very different politicians being elected.

Google was sneaky. They didn't tear up all the voter registration cards at once. Rather each day for the last year Google went door to door and tore up a few voter registration cards. In the constantly running election of Google serps nobody noticed the small changes as they happened each day.

We've been in the middle of a slow boil update. Nobody jumped out of the water or even noticed the fundamental change because it happened slowly over months. Just one observer's opinion.

decaff




msg:3305177
 8:57 pm on Apr 7, 2007 (gmt 0)

Nice Observations..

Certainly with the Supp Index growing .. Google can then calculate per each site's volume of pages in the supplemental listings vs. what they have in the live organics .. and work this ratio as a set of filters ... so if a site has a strong presence in the live organic listings and "some" supplementals ...then this site is probably going to continue to perform well..but if the ratio flips...then filters kick in and the site suffers...

Still, I can see how the comment is open to various interpretations as written. Does it mean don't change the nav menu's anchor text when it appears on different pages? ...or does it mean that even in one page's version of the nav menu, we should be wary of repeating the same keyword in multiple phrases?

Certainly from the "human visitor" perspective .. retaining continuity throughout your site from page to page in regards to how you name your links (internal navigation anchor text) is going to be critical to giving your visitors easy and familiar access to your most important pages..

If you are trying to game the algo with internal anchor text variations...this really doesn't make any sense...as from one page to the next your link descriptions will change..?

Where the true anchor text variation would be expected is with the external inbound links.(from sources pointing to your site)....especially if these have been acquired organically...

If you try to game the algo with an anchor text variation naming scheme for your external inbound link campaigns...this might be spotted as too predicable in nature...

66sore




msg:3305181
 9:08 pm on Apr 7, 2007 (gmt 0)

Google search results seem almost randomized now. There is way too much emphasis on social bookmarking sites, which are being spammed to death btw.

My feeder site is ranking higher for my phrases than the actual site. It's the same when social bookmarking sites and directories rank higher for a phrase than the actual site these directories/social bookmarks point to.

The most fundamental change at Google Seach is that the results could be a lot better!

Robert Charlton




msg:3305189
 9:25 pm on Apr 7, 2007 (gmt 0)

...but varying anchor text seems at the top of the risk list right now (and follows the years of WebmasterWorld "vary your anchor text" threads

steveb - Are you saying that varying your anchor text can hurt you... or simply that it isn't producing the number of rankings that it used to?

At the risk of going too far off topic here...

To look at the latter... It used to be that you could rank for varied modifiers, eg, by including them in your inbound anchors. Now, as Google is tying inbound links more strictly to onpage content, you don't rank for these modifiers unless they're also on your target page. I'm not sure, though, these anchor text variations specifically hurt you. Conceivably, they could still be helping your core phrases by lending "naturalness" to your links, but perhaps only to a degree....

In the phrase based system, though, I do see where excessive onpage variation could help you, but I'd be hard put to come up with that many variants on a page.

I also remember reading something about excessive variation in inbound anchor text too, but I can't lay my hands on it right now. Maybe someone else can find it. If Google is covering onpage patterns in their patents, it's only natural that they'd cover linking.

Here's what Google says about the excessive variation for onpage, with my emphasis added....

Detecting spam documents in a phrase based information retrieval system [appft1.uspto.gov]

[0010] The information retrieval system is adapted to identify a spam document based on the appearance of excessive number of related phrases in the document....  

"[0213] From the foregoing, the number of the related phrases present in a given document will be known. A normal, non-spam document will generally have a relatively limited number of related phrases, typically on the order of between 8 and 20, depending on the document collection. By contrast, a spam document will have an excessive number of related phrases, for example on the order of between 100 and 1000 related phrases. Thus, the present invention takes advantage of this discovery by identifying as spam documents those documents that have a statistically significant deviation in the number of related phrases relative to an expected number of related phrases for documents in the document collection.

Google does somewhere discuss linking patterns that are out of the norm... I'm sure meaning again, too much variation.

[edited by: Robert_Charlton at 9:32 pm (utc) on April 7, 2007]

steveb




msg:3305194
 9:35 pm on Apr 7, 2007 (gmt 0)

"Are you saying that varying your anchor text can hurt you... "

This has always been the case. Wildly unnatural linking has never been a good thing. Additionally the tactic of using linking to try and make unnatural linking appear natural is a spam tactic Google has been aware of for years now.

So it should be no surprise that naturally varied anchor text might mistakenly get you in trouble when the spam scoring is looking for unnatural link patterns.

While this is an issue, I personally wouldn't spend two seconds worrying about it. Linking naturally coincidentally normally works best for search engines and users. Sometimes Google misinterprets signals, but no matter what you do they will sometimes misinterpret signals (sometimes lowering scores and sometimes raising scores, for incorrect reasons).

Marcia




msg:3305203
 9:41 pm on Apr 7, 2007 (gmt 0)

There was a time when it became evident that an excessively high percentage of identical anchor text in inbound links caused a penalty. This was not just casual observation (or guesswork) - it was tested.

Robert Charlton




msg:3305222
 10:12 pm on Apr 7, 2007 (gmt 0)

"Are you saying that varying your anchor text can hurt you... "

This has always been the case. Wildly unnatural linking has never been a good thing. Additionally the tactic of using linking to try and make unnatural linking appear natural is a spam tactic Google has been aware of for years now.

I completely agree with you. You just seemed to be emphasizing the danger rather specifically and dramatically, and I was wondering whether you had some new observations in mind....

Google still likes one way blog spam, but varying anchor text seems at the top of the risk list right now (and follows the years of WebmasterWorld "vary your anchor text" threads).

Here's the reference to the relevant Google patent, which I'd posted on this thread about changing link text [webmasterworld.com].

Here's the section from the Google patent...

Information retrieval based on historical data [appft1.uspto.gov]

My emphasis:
[0119] According to an implementation consistent with the principles of the invention, information regarding unique words, bigrams, and phrases in anchor text may be used to generate (or alter) a score associated with a document. For example, search engine 125 may monitor web (or link) graphs and their behavior over time and use this information for scoring, spam detection, or other purposes. Naturally developed web graphs typically involve independent decisions. Synthetically generated web graphs, which are usually indicative of an intent to spam, are based on coordinated decisions, causing the profile of growth in anchor words/bigrams/phrases to likely be relatively spiky.

outland88




msg:3305356
 2:44 am on Apr 8, 2007 (gmt 0)

Excellent post Rekitty. I'm amazed nobody has brought up the issue and the repercussions of placing so many pages in the supplemental index. Actually by doing so Google solidifies the rankings of many authority sites and destabilizes the rankings of what tend to be smaller business sites.

Just last week two major news services were recanting Google’s mantra of them wanting to index all the world’s information. The reality though seems quite the opposite. Basically except for their chosen sites they seem to have pronounced everybody a spammer.

marbles88




msg:3305403
 4:45 am on Apr 8, 2007 (gmt 0)

To isolate one factor, I don't think that Wikipedia's rapid increase is that big of a mystery.

A few weeks ago they put rel='nofollow' on all of their outbound links. Then, a couple weeks later, we started, for the first time, to see them rank up. Many Wiki pages went to the top of the rankings after that, more because of site rank than individual page rank. If you look at their inbound links on a given page versus others in the results, they're not always that great. It sounds to me like Wiki was bleeding away pagerank at both the site level and also the page level prior to their change in policy.

This was, in effect, a highly controlled experiment in the weighting factor of outbound links.

annej




msg:3305427
 5:50 am on Apr 8, 2007 (gmt 0)

I hadn't put that together but now that I think about it there is a correlation between their deciding to use all nofollow and their rapid gains in the serps. There could be a relationship.

Marcia




msg:3305428
 6:01 am on Apr 8, 2007 (gmt 0)

Naturally developed web graphs typically involve independent decisions. Synthetically generated web graphs, which are usually indicative of an intent to spam, are based on coordinated decisions, causing the profile of growth in anchor words/bigrams/phrases to likely be relatively spiky.

Exactly the point. Over a certain percentage of identical anchor text on inbound links cannot be done with independent decisions. Not 80-90% of the people linking will use the same exact wording of "money keywords" in the anchor of links, it has to be contrived. Natural development would be indicated by a certain amount of variation.

At the time the testing was done, there was a very specific figure - under nn% of indentical anchor text = no penalty. Having nn% or above identical = penalty. I remember the exact figures.

Changing anchor text of existing links is a different matter, and so is internal anchor text in global navigation and otherwise.

JudgeJeffries




msg:3305466
 8:29 am on Apr 8, 2007 (gmt 0)

So how come many sites that had massive reciprocal links campaigns with a set number of variant anchor text are still ranking well. In my sphere the current front runners all had huge campaigns lasting a very long time and now, a couple of years later seem to be reaping the benefits. Nothing natural about any of them.

Crush




msg:3305467
 8:46 am on Apr 8, 2007 (gmt 0)

we listened to the twaddle aboout varied anchor and had silly links with keywords for LSI for a while

I can tell you excessive anchor works. If a page is about red widgests the anchor is red widgets. I know people now pointing millions of links with two word anchor and it works but for that serps only.

If you wants to rank for widgets red as well then you need that anchor. Simple

JudgeJeffries




msg:3305473
 9:03 am on Apr 8, 2007 (gmt 0)

Crush - that is exactly my experience.
It needs to be looked at page by page and variation on an individual page basis is not necessary. I think all this talk about varying anchor text to look natural is twaddle put about by the search engines to diminish the demand for the links companies who have such an obvious effect on the serps.
What they may have done is downgraded the effect of reciprocal links in favour of one ways, and they may have introduced a delay factor but that of course is a completely different matter.
Notwithstanding that possibility however you will still trump a competitor if you have thousands of reciprocal links all saying red widgets, natural or not. Ask some of the spammers who read here how they do it and to a man they seem to see little point in the finesse of varying anchor text. They hit the problem with both barrels i.e. red widget - red widget.
I have recently analysed all of the links to one of my competitors sites. He is a well known spammer who moves in the right(wrong?) circles and every one of his links uses the same anchor text - he beats me every time.

ALbino




msg:3305792
 8:44 pm on Apr 8, 2007 (gmt 0)

What if people are linking to a page about an author? Wouldn't it make sense for 80%+ of the inbound links to be "FirstName LastName"?

glengara




msg:3305795
 8:48 pm on Apr 8, 2007 (gmt 0)

Can't say I've spotted any "Fundamental Change", and anything mentioned so far seems incidental rather than fundamental...

Marcia




msg:3305804
 9:00 pm on Apr 8, 2007 (gmt 0)

Author name definitely, it stands to reason. But how about "prepaid dog walking services Palo Alto" - what are the chances of 80% being like that if the webmaster didn't request (or buy) with specific linking instructions. Dunno, maybe it is working.

crobb305




msg:3305805
 9:05 pm on Apr 8, 2007 (gmt 0)

I know I am behind the times. I haven't even studied this so-called 950 penalty. I have no idea what it is. So if there are fundamental changes afoot, I need to get with the program. Maybe someone here would be willing to summarize the 6 different threads about the "950 penalty" and give us a crash course?

[edited by: crobb305 at 9:06 pm (utc) on April 8, 2007]

ALbino




msg:3305806
 9:06 pm on Apr 8, 2007 (gmt 0)

Author name definitely, it stands to reason. But how about "prepaid dog walking services Palo Alto" - what are the chances of 80% being like that if the webmaster didn't request (or buy) with specific linking instructions. Dunno, maybe it is working.

Well, that's definitely not natural, you got me there.

JudgeJeffries




msg:3305832
 9:55 pm on Apr 8, 2007 (gmt 0)

If you sell 'Flash Bang Wallop Cola' what are most of your links going to say? If there has been a change in ranking I doubt very much if its to do with anchor text and if it is coca-cola.com would be history (and its not!). -950? .....back to the drawing board I think.

Marcia




msg:3305846
 10:11 pm on Apr 8, 2007 (gmt 0)

I feel like a broken record stuck in a groove, but I still can't shake the idea that a lot of what's going on has to do with clustering algos - including *some* of the -950 stuff. Filtered out (down) by clustering, yes - penalty, no. Not all, some.

tedster




msg:3305883
 11:21 pm on Apr 8, 2007 (gmt 0)

Thinking about backlinks - both Matt and Vanessa have confirmed that there was a change this year in the way backlinks are weighted. So that's one effect to take into account with the current SERPs.

I haven't been able to isolate a simple change here - nothing appears to have simply "stopped working". I'm thinking it's more likely to be something with types of links and how they can be combined into a link profile for a given domain.

europeforvisitors




msg:3305927
 12:54 am on Apr 9, 2007 (gmt 0)

If you sell 'Flash Bang Wallop Cola' what are most of your links going to say? If there has been a change in ranking I doubt very much if its to do with anchor text and if it is coca-cola.com would be history (and its not!).

Maybe Google has become smart enough to understand that 100 links with the anchor text "Flash Bang Wallop Cola" are likely to be manipulative unless they're pointing to flashbangwallopcola.com. :-)

jd01




msg:3305932
 1:12 am on Apr 9, 2007 (gmt 0)

Maybe Google has become smart enough to understand that 100 links with the anchor text "Flash Bang Wallop Cola" are likely to be manipulative unless they're pointing to flashbangwallopcola.com.

…and to take EFV's thought a step further (if I was a speculator) I might speculate: if the historical data for links pointing to a site (not flashbangwallopcola.com) indicated there were 10 links added for exactly "Flash Bang Wallop Cola" over a 3 month period of time, along with 50 other links not containing the exact link text "Flash Bang Wallop Cola", but during the following 3 month period of time 100 links containing exactly "Flash Bang Wallop Cola" were added with 50 other links not containing the exact link text "Flash Bang Wallop Cola", then for another 3 month period of time 10 links containing exactly "Flash Bang Wallop Cola" and 50 other links not containing exactly "Flash Bang Wallop Cola" were added, it might appear someone was trying to manipulate their link count/text for a period of time.

That's really just 'far-fetched' Easter speculation on my part and I have no historical data or case studies to back it up, but my guess is it's a little more complicated than, "Oops! too many of the same links / link text", and probably has more to do with patterns of link / link text addition over a period of time.

<added>
Sorry, didn't read all the way back to the original post, just jumped in for a minute when I read some of the latter posts.
</added>

Justin

[edited by: jd01 at 1:29 am (utc) on April 9, 2007]

mfishy




msg:3305933
 1:18 am on Apr 9, 2007 (gmt 0)

Inbound links should be varied. Varying internal linking in your navigation is obviously not natural and should be avoided. Within content, it can be fine, but the original post is about nav links.

bouncybunny




msg:3305956
 2:11 am on Apr 9, 2007 (gmt 0)

When you think about what is natural it is natural to have the anchor text in your navigation menu the same throughout the site. It is NOT natural that everyone who links to you would use the same anchor text.

But it is also natural to vary the anchor text in internal links. I often link to the same story on my site when referencing it. Depending on the reasons for referencing it, I often use different anchor text. It is the most natural thing in the world.

Sometimes I might link to the "red widgets story" at other times I might link to "that great article on widgets". Are Google going to give me a hard time about that? Seems odd.

This 138 message thread spans 5 pages: < < 138 ( 1 2 [3] 4 5 > >
Global Options:
 top home search open messages active posts  
 

Home / Forums Index / Google / Google SEO News and Discussion
rss feed

All trademarks and copyrights held by respective owners. Member comments are owned by the poster.
Home ¦ Free Tools ¦ Terms of Service ¦ Privacy Policy ¦ Report Problem ¦ About ¦ Library ¦ Newsletter
WebmasterWorld is a Developer Shed Community owned by Jim Boykin.
© Webmaster World 1996-2014 all rights reserved