Marketing_Guy - 6:12 pm on May 16, 2011 (gmt 0)
Have a site that previously was getting around 120,000 visitors a month. The site split down to around 300 pages (static HTML) and 24,000 (VB forum content).
Static pages all historically (almost 9 years) ranked very well and post Panda dropped hard (80% loss).
Forum pages saw some reasonable traffic (small percentage of the whole) - post Panda, zero change in traffic levels.
1) HTML site was a poor design. In terms of SEO, I'd rate it as mid to high optimisation (it could be better optimised, but I don't like pushing the boat out too far). I'm certain the design issues are the reason for the loss on traffic post Panda.
2) The HTML site didn't particularly push forum registrations - a few links here and there. Could have been better promoted internally.
3) Forum registrations saw no drop post Panda which suggests the bulk come from forum referrals and not HTML site referrals.
I'd conclude that Panda hit a selective portion of the site based on certain criteria which has left the forum unaffected. That's not to say forums aren't Panda proof - just in this case there wasn't a site wide filter applied.
(Oh and my forums have a strict spam filter and anti link drop policy. Not saying that this is definitely a factor, but just throwing it in there as something Google might look at when determining quality signals in forums).