Search within search? Weird google URLs

[Ericc22]

Good morning afternoon, how are you guys doing today?

I'm experiencing a few Panda issues I'm trying to fix, and I was hoping I could get some help here about one of my problems.

I used Google analytics to extract pages people land on after a Google search. I'm trying to identify thin pages that potentially harm my website as a whole.

It turns out I have a bunch of pages in the likes of the following: /search?cd=15&hl=en&ct=clnk&gl=uk&source=www.google .co.uk, and so on for a bunch of countries (.fi, .com, .sg, .pk, and so on, maybe 50 of them)

My question is: what are those pages? their stats are awful, usually 1 visitor, 100% bounce rate, and 0 links. Do you think they can explain my dramatic drop in traffic following Panda?

If so, what should I do with them? NOINDEX? Deletion? What would you suggest?

I also have a lot of links in the likes of the following:

/google-search?cx=partner-pub-6553421918056260:armz8yts3ql&cof=FORID:10&ie=ISO-8859-1&sa=Search&siteurl=www.mysite.com/content/article

They lead to custom search pages. What should I do with them?

Almost two weeks ago, Dr. Pete posted an article untitled Fat Panda and Thin Content in which he deals with "search within search" and how they might be targeted by Panda. Do you think this is the issue I'm facing?

Any suggestion/help would be much appreciated!

Thanks a lot and have a great day

[adrianvender1]

Looks like your GA code is running when people are viewing the Google cache of your page.  Do a search for anything that you rank for and click on the 'Cached' link.  When viewing the cached page you'll notice that the query string parameters are similar to the first example you gave.

You can set a filter to Include ONLY traffic coming from your domain OR Exclude traffic from specific domains like webcache.googleusercontent.com.

You can also add a setDomainName() method to your GA tracking code and set it to 'domain.com' except use your actual domain.  This will allow GA to ONLY run when the domain in browser contains what you set in setDomainName.  GA is a 1st party cookie solution so it won't run on domains that conflict with what you set in setDomainName.

[gmellak]

Were you able to check if any site is linking to those urls containing /search/ or /google-search/? If so, try to integrate a 301 redirect via htaccess or PHP (shout if you need more info on that).

In the robots.txt, put

User-agent: *
Disallow: /search/
Disallow: /google-search/

to exclude these "directories" from crawling ...

[Ericc22]

Hi Gmellak, thanks for the reply!

When I type www.yourdomain.com/google-search, I basically have a blank page... I'm not sure if this what you mean by "strange"

Moreover, how do i exclude /search and /google-search via robots.txt?

Thanks again!

[gmellak]

I'm not quite sure how these results got into your landing pages list ... is /search and /google-search the usual path you are using on your site? When you type www.yourdomain.com/google-search, do you get somewhere on your page? Maybe some spammer or robot is creating random links towards these URLs and your site responds with a 200 http code and "strange" content. Try to find if there are any backlinks to the URLs people land on and this way find out if someone is linking to your site with these URLs.

You can always exclude /search and /google-search via robots.txt so they don't get crawled, and meanwhile forward via 301 those visits to the page you find most suitable and avoid people really landing on those pages.