Forum Moderators: Robert Charlton & goodroi
I don't think this is related to reciprocal linking. If it is overdone to the point that Google sees a big red flag some other penalty might click in but not this 950 thing.
Phase based seems a lot more likely. And there may be something about the words or phrases used in internal linking involved as well. Or that could just be a part of the phrase based thing.
[edited by: tedster at 9:15 pm (utc) on Feb. 27, 2008]
The landing page is swimming pool lights, no gambling. The campaign is up and running after a manual review.
It shows that ad words flags certain terms; I would say the algo does as well. Google will put their resources of manual review for ad words since it’s $$$. For the algo there is probably not as many manual reviewers.
Funny thing is, we rank fine (top 10) for the term in Google main. We were setting up a content network advertising campaign.
A few hundred messages ago I wrote about how I decided to shut down a page that was 950ed. So I striped it of all content, everything including title and metas. It was totally blank except for a couple of one word links. The page came back in just a few days.
I am convinced that it came back because the suspect word or phrase was no longer on the page. I put the page back online but without words and phrases I thought might have been the problem and the page continues to be up at the top of the search results even after a couple of months.
So a lot of the analysis of other factors many be wasting our time. It's easy to assume the 950 was caused by something more obvious as it's very difficult to discern words or phrases that are triggering this.
I know the next thing that you will wonder is why some pages with the word don't get the filter applied. That's where strength of links from related sites come in as well as the fact you might have some other words or phrases that negate the filter. But just try to guess what those are. Usually it's about impossible.
Our lost keyphrase is 1st key + 2nd key (1st key is in domain, 2nd key is same as pagename)
1. Putting our 950 url in the keyword suggestion tool shows only 2 suggestions
! Both suggestions are: some ones name + 1st key (some ones name is not on the page)
2. Uploading the same page with an other filename (2nd key+tmp.html) also shows only 2 suggestions! Both suggestions are again the same: some ones name + 1st key
3. Uploading the same page with a different filename (0.html) shows a huge list of suggestions related to the 1st key
4. For every other page not in the 950’s the tool shows a huge list of related suggestions but only for the 1st key
5. Making the original 950 url return a 404 shows a huge list, all related to the 1st key.
For every test the stats indicated those files were real time accessed by adwords bot 72.14.194.31 …… "cks"
The only thing that comes in mind is pagename = keyword, but only one of our pages is 950!
Any other thoughts on this?
The problem with multiple variables is that it is too easy to refute one in a specific example, but still have it matter.
I'm able to see end of serps sites return the same quality of keywords on the adwords tool as their non-reranked peers, both of which have similar use of on-page factors (title, meta, h tags, etc).
Given that you are seeing a difference in your example may point to your specific problem, which I hope is the case, so you can pinpoint and correct whatever flavor-of-the-week problem Google has with your site.
Cygnus
In the case of the -950 phenomenon, there are most likely several possible triggers for one "penalty" mechanism - and each trigger would have its own threshold that Google can dial up or down.
Adam Lasnik: As algorithms get adjusted, most sites are not affected. There is not a huge shift for most sites, but then some sites happen to be right on that border. I sympathize and I understand that it can be painful for those sites that are on a particular threshold on one or more algorithms.[stonetemple.com...]
I didn't say anything about anchor text.
Google has long had various link power penalties... too much too soon being one talked about endlessly.
There are largely two types of pages 950ed: those that deserve it, and those that don't. A problem with these threads is some people are talking about how to trick Google to prevent them from recognizing and punishing garbage/spam. I hope all the pages Google is genuinely trying to 950 penalize stay penalized. I personally couldn't care less about that. The misapplied penalties are what concern me. These tend to effect (fairly) powerful, quality pages on authoritative domains.
Its oversimplified to say Google is trying to detect false power and is collateral damaging genuine power at an alarmingly inept rate... but that is kinda the basic issue.
These tend to effect (fairly) powerful, quality pages on authoritative domains.
Exactly - those false positives are taking away some very good results.
I removed the sitemap after reading a thread somewhere just to see if it had any effect. I'm not saying that this is the solution. Other than that, no major changes to site - I didn't want to make things worse. So I don't know why I'm at #1 again and when I will go MIA again...
I also noticed that the serps for another major keyword has reverted back to May 01. Anyone notice the same for their category?
[edited by: andrewshim at 10:18 pm (utc) on May 8, 2007]
But obviously the presumption of innocence doesn't work with Google, it's more like a presumption of guilt.
[edited by: SEOPTI at 10:29 pm (utc) on May 8, 2007]
Its oversimplified to say Google is trying to detect false power and is collateral damaging genuine power at an alarmingly inept rate... but that is kinda the basic issue.
Well we've been in agreement on this one I guess.
I've been a little worried about actually whom I am talking to trying to "lift the penalty". I was thinking of genuine websites, but... You may be right, this information could be used by anyone.
I was thinking of genuine websites, but... You may be right, this information could be used by anyone.
It could be used by anyone but I doubt anyone who is spewing out thousands upon thousands of automated pages is going to bother searching through each page, analyzing what words or phrases might be the problem. It just wouldn't be cost effective for them. So some of their pages don't work, it's no big deal as others will.
If an established site has backlinks to it from other authority sites chances are that its unlikely to be a spammers site.
The PR of a site and the quality of the backlinks to that site should count for something. A low quality site with little content designed for spamming reasons is unlikely to attract high PR backlinks, .gov backlinks, .edu backlinks etc etc yet this 950+ penalty is hitting the very sites that add quality to the index.
Currently its a Joke and the biggest laugh is that some of the spammiest sites known to man still rank in the dam index whilst a hign volume of authority sites are in the 950+ club. You couldnt make it up!
Don't use analytics, if you do you will see your sites drop 950+ in many cases.
Reminds me that I got a friendly invitation to adwords before my traffic stopped. A friend did my adwords. Hey they didn't get that yet... Surveillance isn't complete yet .. Well we obviously pulled out of adwords ...
1 one-way inbound link
1 one-way outbound link
6 reciprocal links
Showing up with rankings fluctuating between high 20's and high 30's for the primary keyword phrase, for the first time, with only those few links is quite OK in my book.
The landing page is swimming pool lights, no gambling. The campaign is up and running after a manual review.It shows that ad words flags certain terms;
I would say the algo does as well. Google will put their resources of manual review for ad words since it’s $$$. For the algo there is probably not as many manual reviewers.
So then, are you saying that Google search is primarily based on human editing review and standards, just like a directory is, rather than on algorithms? Please provide us with evidence of such a HUGE paradigm shift from using algorithms to using human reviewed directory-type inclusion, if you kindly would.
If there's been a major difference in how search engines work, surely it needs to be discussed So please tell us how it's all changed from crawling, indexing and algothmically ranking by relevancy factors.
AFAIK the perception that the SERPs are human edited rather than algorithmically sorted, if anyone has any evidence, should most certainly be documented and/or discussed.
Funny thing is, we rank fine (top 10) for the term in Google main. We were setting up a content network advertising campaign.
Remmber with the phrase based thing just one word or phrase can put you in the 950 regions.
Ever think about why?
C'mon folks, tell us about information gain. ;)
Posted by trinorthlighting:
It shows that ad words flags certain terms; I would say the algo does as well. Google will put their resources of manual review for ad words since it’s $$$. For the algo there is probably not as many manual reviewers.
So sorry compadre, but with all due respect and with sincere apologies, that is the biggest crock of horsehockey I've read in a month of Sundays.
[edited by: Marcia at 8:36 am (utc) on May 9, 2007]
Remember with the phrase based thing just one word or phrase can put you in the 950 regions.
One added word tacked on to a "good phrase" might not give the same potential for information gain, possibly because of contextual analysis and/or co-occurrence statistics (more than likely the latter), and send an otherwise good page to the nether regions of purgatory.
Reminds me that I got a friendly invitation to adwords before my traffic stopped.
Hmm. Seems I might not be the only one who thinks these two events might be related.
Oh my gosh! Help me please, I'm emotionally trembling.
I've just recently received offers in the mail for burial insurance and pre-paid cremation services. Does that mean I'm about to die?
What should I do?
I've just recently received offers in the mail for burial insurance and pre-paid cremation services. Does that mean I'm about to die?What should I do?
If I were you, I'd sign up immediately. Much cheaper, and who knows, you may not get enough Google traffic to provide the income to foot the bill at a later date. ;)
I suppose the adwords team or even adwords user have the potential to flag a site which may be interpreted incorrect by their algorithm.
So does that mean that Matt Cutts and all of the search engineering team are under the control/domination/direction of the AdWords CSR's?
That's amazing! It certainly is far more power than any of us who worked customer service/sales/marketing in the B&M world ever had over any of the others who worked the front lines.
So then, is it safe to assume that Matt and his whole spam_team and quality control people and the search engineers are nothing more than puppets, moving and adjusting search algorithms at the direction of Adwords CSRs based on bids and market share?
I thought the search folks had more autonomy than that, but maybe I was wrong Maybe it's the AdWords CSR's who run search after all. Wouldn't that be a kick in the head, if all of us have been wrong all this time.
I've just recently received offers in the mail for burial insurance and pre-paid cremation services. Does that mean I'm about to die?What should I do?
Well i also got a Google friends newsletter... :)
Google seems rather compartmentalised and I have a feeling their internal communication might be not very good. If it would be smashing and the adwords folks would be best pals with the SE things, might be, well then we have some form of manipulation (in a subconscious way), possibly. Between tin foil hat and total trust there are many shades of grey.
And there is always the possibiliy of a bug ... huge masses of data controlled by comparatively simple algorithms will always be prone to unexpected results. Especially if you increase the dynamics of the system (aka Big Daddy) see chaos and so on. Must be hell of a system to keep tabs on. Gazzillions of data, gazzillions of computers ... There must be unexpected effects all the time.
[edited by: mattg3 at 12:23 pm (utc) on May 9, 2007]
