Forum Moderators: Robert Charlton & goodroi
Any suggestions on what to do?
Thanks Dave
Here's a thread that's worth a read:
Sudden Drops in Rank - a checklist [webmasterworld.com]
It took me about 3 months to finally discover what happened.
SO let me save you the time, hopefully:
Duplicate Content.
I am willing to bet, and I think I will win, that you have a duplicate content problem somewhere at your site.
When it happened to me, I ignored the possibility, did cursory checks, and always decided I didn't have this problem.
But then I discovered I did!
It took me months to find it, some time to fix it, and in the end, all is well again, but man, the end rsult is that 2006 was a bad year for us.
So, please, make sure you do not have a duplicate content problem. check:
- #1 suspect: many different variations of your URLs actually pointing to the same page. I bet this is yoru problem! FIX IT NOW! You will find your rankings come back in 2-6 weeks. i.e. mysite/param1#*$!/param2yyy vs. mysite/param2yyy/param1#*$! - different urls, but goign to the same page, you will get massacred at google.
- www.mysite.com vs. mysite.com, called "canonical" - it will kill you, but it wasn't my problem
Do not leave any stone unturned at your site. I bet this is your problem. Do not quickly assume it is NOT your problem. I did that, and it cost me 6 weeks of lost revenue - I stupidly sat cheerily on duplicate content issues at my site until I finally admitted to myself I am not perfect, and maybe I did have such a problem, and I should look again, and lo and behold, there it was.
I truly hope I just saved you some time and grief.
last night I checked google still #1, this morning we are #46 and it is not our index page that would comes up.
With the rolling updates, I have found similar situations that eventually straighten out in a day or 2. I think it might all be in the timing. A well established site that typically garners #1 for particular phrases can tank as data refreshes. I would only worry if it stays at its present postiton for more then a week. Then you got problems...
- #1 suspect: many different variations of your URLs actually pointing to the same page. I bet this is yoru problem! FIX IT NOW! You will find your rankings come back in 2-6 weeks. i.e. mysite/param1#*$!/param2yyy vs. mysite/param2yyy/param1#*$! - different urls, but goign to the same page, you will get massacred at google.
I hope it's not true.
My understanding is that in such a case Google picks up one page as the original one and ignores the rest, so basically there shouldn't be a problem with it, if all the similar pages belong to the same site.
I could only imagine a glitch in the Google's algo where the mere presence of a duplicate page, regardless of original being on the same site, downgrades the whole site in general, but I am very doubtful about it.
I have links which point to same page where the Anchor text may change. for eg
<a href="pageA.html">WidgetA</a>
<a href="pageA.html">WidgetALongName</a>
Will this count as dup content? Did u have to clean somthing like this?
Also what if I have
<a href="/products/widgetA.html">My Product</a>
<a href="/index.html?id=abcd">My Product</a>
<a href="/products/widgetA.html">My Product LongName</a>
<a href="/index.html?id=abcd">My Product LongName</a>
All pointing to the same page. Would this hurt? I guess it would.
Also, isnt www.xyz.com and xyz.com be fixed as same from google webmaster tools site?
Please let me know. Thanks
Leo.
A well established site that typically garners #1 for particular phrases can tank as data refreshes.
We are seeing this quite a bit, long time high ranking sites suddenly vanishing on some data centers. Very scary to say the least but after 4 or 5 days everything seems to come back to normal. Best way we have found to identify this is to search for the same site using another key word, the second or third most popular. If you find the site in its usual good ranking position for these secondary terms, most likely its one of these data pushes, refreshes or whatever it is.
(still cannot understand why, or even how, a site is temporarily removed from a data center, just for a particular key word)
In your first example, the pages are the same. It is simply the anchor text or link text that is different. That is fine.
In your second example, you list 4 different URLs pointing to the exact same page. That IS a problem. You only want 1 URL pointing to the page. Multiple URLs pointing to the exact same thing confuse the hell out of Google and often lead to the duplicate content issues.
In your third example you have the www and non-www versions. In a perfect world, these shouldn't cause problems, but they do. The webmaster tools site let's you tell G which one you want used so this is a definite step forward, but you still should also work to solve the issue on your end (redirect non-www to www or visa versa).
[edited by: Philosopher at 7:47 pm (utc) on Nov. 20, 2006]
I recently launched a new site where mydomain.com simply did not exist anywhere, never had ever but Google CREATED it within days. If this is a problem for Google why should it have created this very problem?
Anyone have an logically support answer I would love to hear it.
It's been discussed many times, but the simple fact is that www.mydomain.com and mydomain.com are technically different sites even though in most circumstances, they both point to the same place.
As for why Google would pick up a version you aren't using...
1) Outside links pointing to the wrong version
2) Internal links using the incorrect version
3) Site owner, or anyone else with toolbar installed typing in the wrong version when visiting (I do that all the time and google DOES pick up URLs visited via the toolbar and index them).
lots more.
Once G has both versions the trouble begins.
also, does having the mysite.com not work at the IIS level or server level fix this problem?
Thanks
Leo
Once G has both versions the trouble begins.
I have links which point to same page where the Anchor text may change. for eg
<a href="pageA.html">WidgetA</a>
<a href="pageA.html">WidgetALongName</a>Will this count as dup content? Did u have to clean somthing like this?
no, not at all unless you do not overuse it IMO.
But it is true.
The same content at multiple different URLs - even if the URL differs by just one character - is duplicate content.
I think its dangerous and confuses google as to what the page really is about. Your anchor should be the best description of the page, not one of many
I never risk it, especially with the two anchors on the one page.
The same content at multiple different URLs - even if the URL differs by just one character - is duplicate content.
I didn't question that.
The point is that ONE of the URLs is still the canonical one, properly indexed and not filtered.
So, one shouldn't bother about duplicates as they are where they belong to.
The question is whether the simple occurence of duplicate pages on the site, cause the filtering of the canonical one too, as some posts imply.
Additionally, as the PageRank and both the internal and external linking changes, you will find that they revise the choice from time to time.
That is how you get a site half listed as www and half listed as non-www, and with 90% of the URLs tagged as Supplemental ... and with a little bit of thought it need never have happened that way at all.
As g1smd stated, even if it doesn't happen to cause a dup. content issue, there are LOTS other problems it can and does cause.
Ah, but the point is that Google chooses which one to list (and it might not be the one that you wanted to list), and then filters the rest (usually drops most copies into Supplemental).
Right, but that is a different animal, altough extremely important.
Lets say someone has most if not all valuable links pointed to www.domain.tld/various.html's and suddenly due to some unpredictable linking or other cause, SE picks domain.tld/ (non-www) as the canonical root url.
The chance is huge that ALL the previously well ranked pages will go supplemental just because of the SINGLE wrong canonical url one level above.
So yes, one should be very aware of WHICH URL's are canonical on his/her site in case of duplicate pages.
However, this still does not answer the question about the possible bad influence of duplicate URL's towards the canonical one and the site in general.
** This is _NOT_ duplicate content. No worries!
NOT duplicate content!
++++++
It WOULD be duplicate content if you had this:
<a href="/index.html?id2=yyzz&ad=abcd">My Product LongName</a>
<a href="/index.html?id=abcd&id2=yyzz">My Product LongName</a>
(Notwithstanding a concurrent discussion about whether or not google will see everythign after the '&'.)
The point is, if at your site, you have a page that can be accessed via DIFFERENT URLs, then you have a problem. Becuase each of those URLs will be considered DIFFERENT pages by google, but when google goes to look at them it will see they are the same DUPLICATE content pages, and that is where the problems start.
Also, in the "good" data center, we are the only result on the top 30 that has an indexed result:
www.example.com
www.example.com/red_widgets.htm
What do you think?
