|Google.com only - Penalty/filter/sandbox?|
| 8:22 am on May 8, 2009 (gmt 0)|
I have an issue that I have not see before. Basically I have been fairly aggressivly building links to my homepage, all 25 targeted keyphrases have tanked, not in top 1000, but only in google.com.
All other traffic, for pages/keyphrases that were not targeted remains fine, across all Google's.
Any ideas what has happened here, a google.com only filter/penalty/sandbox? How can I get out of this?
| 4:41 pm on May 8, 2009 (gmt 0)|
Hey Dave_Hybrid. I don't know if you've given WebmasterWorld enough information. Some ideas for possible things that could have incurred either automatic or manual penalization include:
- You used any of a whole bunch of techniques that Google is known to penalize a page for if they're discovered, including hidden links or text, excessive use of would-be optimized HTML elements (so-called "overoptimization", not a great word), or duplicate/redundant content, which isn't so much a penalty as much as a filtering for "failing to qualify as relevant".
- You say you've been "fairly aggressivly building links". How do you mean? Are you using paid links? Are they obviously paid? How about this: Are the sites with the links themselves incurring any penalties? Definitely review how those sites/pages themselves are doing in Google. That alone might account for everything.
- Not a reason for penalization, but does the site have a Google Webmaster Tools [google.com] account? You'd be amazed at the amount of info you can get there, including, sometimes, direct messages re possible penalties.
I'm not trying to be flip, but I'd highly recommend going through Google's Quality Guidelines [google.com] and systematically ruling out all the potential issues they highlight. When you can rule all that stuff (plus all the above points) out, then WebmasterWorld members might have other ideas for you.
| 7:05 pm on May 8, 2009 (gmt 0)|
Without giving specifics, what sorts of sites have these highly targeted links come from?
Have you tried hitting different DCs to see if it is all of Google.com - could just be the DC you hit. Unlikely due to load balancing etc but if you try to hit a load of different ones you shoudl be able to see if it is all/most of Google.
| 8:48 pm on May 8, 2009 (gmt 0)|
I've been doing this a while and have seen similar things before, assume I am familiar with the webmaster guidelines and WMT.
I have just been doing directory and article submissions, only the homepage is affected, only for the keyphrases I have been targeting and only in google.com from a US IP.
It's a bit odd. I've worked on penalized sites a few times and normally the are blanket full google penalized.
| 8:50 pm on May 8, 2009 (gmt 0)|
That's if it is a penalty at all. Seems likely as I have only dropped on the homepage and for the keyphrases I was building links for.
| 4:35 am on May 9, 2009 (gmt 0)|
Too much backlink anchor text that exactly matches a target phrase can create a ranking problem for the phrase - have you been doing that? Natural backlinks have a natural distribution nof anchor text that begins with the plain-old domain name as the most common.
My sense is that Google measures this natural variation in anchor text, and they will filter/penalize for a keyword phrase if the find a statisically significant deviation from the norm. Such a situation is a sign of intention to manipulate the rankings, rather than the kind of "editorially given" backlinks Google says they prefer to see.
From what I have run across, this kind of algorithm calculation is more commonly seen on google.com over their ccTLDs, just as you've noted.
| 10:34 am on May 9, 2009 (gmt 0)|
Thanks tedster, so if I start using many other variations of anchor text or just the plain old URL that should help...
| 5:04 pm on May 9, 2009 (gmt 0)|
It's certainly worth a try. If it is as I suggested, then this particular filter might be a bit sticky to get out of.