PDA


View Full Version : seo issue with search


tfw2005
09-18-2007, 03:03 AM
I am noticing that when searching, the results show the SEO urls for the entries, with ?catid=searchresults&searchid=XXX at the end of it.

This string also gets applied to the url of every single entry url on the search result page.

This number changes each new search.

Having anything picked up in the engines other than the straight forward regular SEO URL as seen on normal Browse Cat pages can trigger duplicate content penalties.

FOR BROWSECAT ISSUES

I have added a conditional to the browsecat template for now, which should address the issue partially.

<if condition="$_REQUEST['do'] == 'searchresults'">
<meta name="robots" content="noindex,nofollow">
</if>


This tells the engines not to index the search result pages, since the searchid can be entered falsely 1 - 1000000 and get the same page.

Since that same string is applied to all entries on the results page, we have to tell it not to follow those URLS either, a dead end for spidering, which sucks.

If you could get the entry URLS to appear on result pages without the string attached, you could do the following:

[code]<if condition="$_REQUEST['do'] == 'searchresults'">
<meta name="robots" content="noindex,follow">
</if>


That way, dynamic url based pages wont get picked up. But the links on them can be followed, and indexed. Currently, cant tell them to follow because of the potential dup content issue.

A completely seperate single SEO command for searches would also fix this, such as http://www.url.com/dyna/search/keyword-keywords - with no variables, and allow the search pages themselves to be relevant non dup content (somewhat, as much of a risk as multi-level categories, which seem to be handled fine by engines) (similar to what I was saying to do with the tags situation in another thread, I would prefer these be the same - tags/search = one).


FOR SHOWENTRY ISSUES

Since direct urls to entry pages with the search string can be copy/pasted to other websites, forums, etc, that could become a spidering point for the engines. If it has the string in it, it would be indexed, and eventually registered as dup content, even tho you nofollow-ed it on your own site. So, I put this there:

<if condition="$_REQUEST['do'] == 'searchresults'">
<meta name="robots" content="noindex,follow">
</if>


This tells engines not to index that page, but follow the other pages. Seems like all the others appearing there are good to go. The dynamic based previous/next results go to the next result, but that page loses the search string. I think thats an error on your part, but the way I want it :).

This sucks because the page it first found you are telling it to ignore, losing relevance if the link on the other site was related to your topic.



Dont want to have this great SEO based URLS thing going on and have it ruined with 3000 searchids pointing to the same results, each of those results with 3000 searchids attached to the end. All it would take is one malicious person to mass copy/paste links with fake search ID strings at the end on a couple different websites, engines see and follows, boom, you are screwed.

I know the variables need to be there for search pagination and probably next/previous situations, but I know Wordpress handles this somehow, so maybe reference them for how they do it.

Brian
09-18-2007, 10:14 AM
I'm not really seeing how this would be an issue, as search engines are obviously not going to be using the search system. Any search results should be deleted after an hour, so any links to those search results pages will result in an 'Invalid Search' message, and thus, no duplicate content.

tfw2005
09-18-2007, 04:51 PM
http://www.vbadvanced.com/dynamics/browsecategory.php?do=searchresults&searchid=91
http://www.vbadvanced.com/dynamics/browsecategory.php?do=searchresults&searchid=92
http://www.vbadvanced.com/dynamics/browsecategory.php?do=searchresults&searchid=93
http://www.vbadvanced.com/dynamics/browsecategory.php?do=searchresults&searchid=94
http://www.vbadvanced.com/dynamics/browsecategory.php?do=searchresults&searchid=95
http://www.vbadvanced.com/dynamics/browsecategory.php?do=searchresults&searchid=9111111

http://www.vbadvanced.com/dynamics/chit-chat-4/multipage-test-20384/
http://www.vbadvanced.com/dynamics/chit-chat-4/multipage-test-20384/?catid=searchresults&searchid=91
http://www.vbadvanced.com/dynamics/chit-chat-4/multipage-test-20384/?catid=searchresults&searchid=92
http://www.vbadvanced.com/dynamics/chit-chat-4/multipage-test-20384/?catid=searchresults&searchid=93
http://www.vbadvanced.com/dynamics/chit-chat-4/multipage-test-20384/?catid=searchresults&searchid=94
http://www.vbadvanced.com/dynamics/chit-chat-4/multipage-test-20384/?catid=searchresults&searchid=95
http://www.vbadvanced.com/dynamics/chit-chat-4/multipage-test-20384/?catid=searchresults&searchid=10000000

http://www.vbadvanced.com/dynamics/sychevs-blog-8/apply-for-beta-testing-20331/
http://www.vbadvanced.com/dynamics/sychevs-blog-8/apply-for-beta-testing-20331/?catid=searchresults&searchid=91
http://www.vbadvanced.com/dynamics/sychevs-blog-8/apply-for-beta-testing-20331/?catid=searchresults&searchid=92
http://www.vbadvanced.com/dynamics/sychevs-blog-8/apply-for-beta-testing-20331/?catid=searchresults&searchid=93
http://www.vbadvanced.com/dynamics/sychevs-blog-8/apply-for-beta-testing-20331/?catid=searchresults&searchid=94
http://www.vbadvanced.com/dynamics/sychevs-blog-8/apply-for-beta-testing-20331/?catid=searchresults&searchid=95
http://www.vbadvanced.com/dynamics/sychevs-blog-8/apply-for-beta-testing-20331/?catid=searchresults&searchid=96
http://www.vbadvanced.com/dynamics/sychevs-blog-8/apply-for-beta-testing-20331/?catid=searchresults&searchid=97
http://www.vbadvanced.com/dynamics/sychevs-blog-8/apply-for-beta-testing-20331/?catid=searchresults&searchid=9110000

Now imagine if that was done maliciously, and posted on dozens or hundreds of websites, and all those individual entry pages are starting points for google spidering.

Also, having the invalid search message is also bad, since someone can copy/paste the url of the search results page into another forum or websites pages. Dead end for the engines.

tfw2005
09-18-2007, 05:00 PM
And another issue, somewhat related:

If you have SEO turned on, and then cut the url down to this manually:

http://www.vbadvanced.com/dynamics/browsecategory.php

It returns your main listing, which is the same as:

http://www.vbadvanced.com/dynamics/header-category-1/

dup content there too.

Brian
09-20-2007, 11:01 AM
Most search engines should ignore the part after the ? in the URL. Pretty much any SEO functions would be left open otherwise. For example, if you're using vBSEO, you could just as easily format links like this:

/forum/1-test.html?highlight=test
/forum/1-test.html?highlight=something
/forum/1-test.html?highlight=nothing

Or for that matter, you can add pretty much whatever you want to the end of most any vBulletin URL and end up with a different URL with the same content just as easily.

If you have SEO turned on, and then cut the url down to this manually:

http://www.vbadvanced.com/dynamics/browsecategory.php

It returns your main listing, which is the same as:

http://www.vbadvanced.com/dynamics/header-category-1/

dup content there too.
That shouldn't count as duplicate content as there is a redirect to the SEO'd URL when you try to use the normal browsecategory link.