Getting Killed by Supplemental Results

My network of sites has recently been punished badly by Google.

I saw most my pages got delegated to supplemental results. Traffic is down by over 50% for the past 3 weeks. With revenues to match

I was using sitewide footer links on most of the sites to point to the other sites in my network. That may have been the problem. I'm removing them now and hope things improve.

And working on getting more deeplinks to internal pages.

Anyone have other suggestions?

 

This typically means you pages do not have enough PageRank to be included in the index. If you pages are not duplicate content you should be able to get out of supplemental hell with some quality back links.

 

No, I don't think PR matters. Some pages in main index are PR0 while some in supplemental are PR3. And all pages are original,wrote them myself.

 

It has nothing to do with PR, but usually is because of the content not being original, low percentage of original content compared to total code, or more recently, running affiliate links on the page.

Crosslinking sites, especially if they are on the same class C IP, isn't a good idea - but I don't think that is what is causing the supplemental results.

 

you think the affiliate links actually cause the supplementals?

 

Here is a not very clear post by Google's Matt Cutts on this topic (read the comments too)

http://www.mattcutts.com/blog/myth-busting-virtual-hosts-vs-dedicated-ip-addresses/

I still can't tell if this is fact or fiction.

 

It certainly can. The vast majority of amazon stores are supplemental (I haven't seen one that isn't, but that doesn't mean there aren't some). Google adwords has just recently hit sites that are running affiliate links in an effort to "improve search results" - so if they are willing to lose money, its reasonable to conclude they have the same policy for natural SERPS which they make nothing off of. Google's position is that affiliate sites are just offering duplicate offerings of an original site - and that visitors are tired of being directed to "affiliate sites". I posted yesterday about this happening with a PR6 site I've had for 8 years. As soon as I masked the affiliate links, SERPS returned (7-10 days).

I don't think every site that runs an affiliate link is going to automatically go supplemental or have SERPS affected, but I think it is now factored in to their algorithm.

 

i just assumed amazon stores are duplicate content, actually all affiliate datafeeds for that matter

it hurts my brain to read the comment on matt cutts blog

 

It seems pretty clear to me:


"I’m happy to affirm that this statement which was true in 2003 is still true now. Links to virtually hosted domains are treated the same as links to domains on dedicated IP addresses."

Google assumes that a virtual host is the same as a dedicated host i.e. links are treated as an owner "voting for their own site." The chances that someone is naturally linking to your site from the same IP is virtually zero. As a registrar, google can also check domain ownership so they can tell if two sites linking to each other have the same ownership. This is also talked about in their patent. If you want to cross link your sites to look like "natural links", use different hosts. There is nothing wrong with putting a link from one site you own to another (even if on the same IP), just don't expect google to treat it as an unbiased "vote" for the site. Putting a bunch of footer links to multiple sites you own starts to look like a link farm trying to manipulate SERPS.

 

Hmmm, this is what I was wondering too. It was primarily my sites with affiliate advertising on them that were harshly affected by this.

How are you masking your links. I'm thinking of pointing the URL's to a fake page on my site and then using a 301 redirect in .htacces to point it to the proper affiliate coded URL. And using a no follow tag in the fake URL. Do you think that is enough? Or should those "fake" pages be placed in a separate folder, and use robots.txt to disallow bots from looking in there, as a double measure.

Should I bother to even build a page for these "fake" pages. An .htacces redirect does the job with or without such a page residing on the sever.

 

My site wasn't even designed as an "affiliate site", but did have 1 - 2 affiliate links on some pages. Only the image had the affiliate link - and the only reason they were there was so I could legally use their product images. I lost all top ten rankings for some very competitive terms that I had ranked well for most of the last 4 years. After I masked the affiliate links, rankings return. There is never 100% that any change was responsible, but imo, this was the cause.

 

Are affiliate links and banner spots the same for returning such a high supplemental count? Just asking a ?..hoping it's not too silly.

 

i have been hit by the same, im wondering if canonical issues come into this, as my site only has a few links , how does having different content in the site affect it eg car site with loan pages in it?

any ideas?

 

That's how I did it. I didn't use the "no follow" based on some things I have heard, but the robots.txt should accomplish the same thing. It works great as long as you don't have too many links (keeps the htaccess file small).

http://forums.digitalpoint.com/showthread.php?t=182984

 

do you think they count adsense as affiliate links?

 

No, adsense or YPN presently are not affected. I'm talking about CJ or Linkshare - but I'm sure it could be others like azoogle, etc.

 

Now, for the dirt - how to get out.
1. If you stole content - change it.
2. If there’s no content - add some.
3. If it’s orphaned - link to it.

ref: http://www.jimboykin.com/damned-to-google-hell-supplemental-results/

 

its java coded so its ignored, i believe.

This could be the one and only positive issue with the new, and largely hated, java based links form CJ.

 

Respectfully, the quote from Matt is not the same as your statement.

He speaks only to dedicated IPs vs. the classic VPS as if the subject domain were the only owned by the publisher.

He never addressed cross linking in the /24 space, but the comment questions did.

And with private registration the whois data may not be available for checking... unless mighty G even has access to that.

Do they?

 

ok done all those

any others?