smithaa02 - 12:52 am on Oct 7, 2012 (gmt 0)
Thanks for the heads up...might have to try that.
Part of the reason I started this thread was that one of our sites got hit in July at around the time of Panda 3.9.
Prior to this update, we had a large number of commercial links that were tracked through an intermediary tracking page + GET variables. In WMT, google was told to treat the tracking link as one.
So the old links looked like:
Tracklink.php would then 302 to the destination page inside of a frame with our logo on the top frame.
What we changed this to, was an openx system from a subdomain. eg
Now everything on adserver.maindomain.com was (in theory) blocked by robots.txt.
But was it a major mistake in switching our outgoing links in this fashion?
Did Panda's machine learning algo correlate bad sites with site that tended to track outgoing links? And therefore openx got flagged?
Or perhaps, linking to an external domain (even if a subdomain) then redirecting 302 to the site is bad?
Perhaps google considered my Tracklink.php as one url (because I used the WMT consolidate variable trick)...and with the change, google thought I was flooding my site with external links all at once?
Perhaps it wasn't even these links?
I know google is touchy about doing mass link updates... Our company once updated all of our say 'section titles' to link to 'special information pages' that already existed. These special information pages did great in the rankings...then tanked.