|Have I been blowing 'link juice' all these years?|
Maybe I'm not as think as I smart I am :-)
I'd like some opinions on this, and I don't recall anyone ever bringing this situation up before, so at least if I've goofed, maybe someone else here can learn from it ...
Every product on my site goes to a third party cart - been doing it for years, and worth the hassle to me to not have to deal with maintenance/security issues.
So, each product essentially has coding something like this...
<form action="http://www.some-third-party-cart.com/somecart/cartthing.cgi" method=POST>
<input name=item value="cart55^MyProductNumber^My Description^9.95>
So, the question is - have I been SCREWING myself all of these years because these are being followed as links in the eyes of the search engines, bleeding off valuable pagerank/linkjuice/etc from virtually every page on my site?
I've tested putting rel=nofollow into that form line, and the cart still works fine. The question, however, is if the search engines will recognize it since it's stuck inside a 'form', and not a standard hyperlink?
Thoughts? Opinions? Hysterical Laughter?
That is not really a link. That is a form submission. I have not tested it but I'm pretty sure Google dose not count that as a link.
When you say "Every product on my site goes to a third party cart" do you mean that the product page is on your site but the buy now button goes to another site? If so I think you have it set up correctly.
Yes Ogletree, the product pages are all on my site - but the buy button goes straight to the shopping cart site (it's .cgi page, at least).
I started thinking about having multiple links like this on every page, and wondered if it was 'hurting' me a bit. Ranking has always (knock on wood) been pretty decent, but if this has been a negative, correcting it would/should provide a little extra 'oomph'.
One thing that may or may matter - the page (and the way we link to it) is an http page (not https)....it does change to https during the actual checkout process.
The .cgi page on the cart's server has a Pagerank of 5, and is showing thousands of links to it in Ask, and tracking those links back, the majority of them seem to be coming from product pages of various merchants.
Yahoo shows the same. Even Google shows a few backlinks to that page which are from another merchants page where the link can only be coming from the url inside the form.
I'm not smart enough to know (or to be able to test) to see if Google and the other search engines actually 'count' those links.
***I assume I can't post the exact URL (and name of Cart Company) here unless a mod decides it's worthwhile as an example to for clarification, but I can sticky it to anyone that wants to look at it further.***
I don't know whether unlinked URL references within functional contexts would affect the flow of PR, but if you're worried about it here's a thought:
Add a few more links to your internal navigation to increase the ratio of internal links vs outbound links. That would change the balance and reduce PR leakage a bit ... if there was any, that is.
I don't know if Google passes link juice on non-linked URL references, but I do know that Google follows said references unless they are blocked by the robots.txt file. My Webmaster Tools Crawl Errors page always contains bad URL references that Google picked up by reading the abridged version of a URL shown in the link text of postings by forum software like vBulletin. When I can figure out where the URL was really supposed to go I will add a 301 redirect for the broken URL to the proper URL.
As said above, making sure to have plenty of internal navigation links on pages will reduce the amount link juice any one outbound link costs you (in theory at least).
If it aint broke don't fix it. Even if it could help you by making a change it is really not going to affect your ranking very much. Some things you can do.
1. as always build more links not only to front page but to internal pages.
2. Make sure your content is unique. If your using descriptions from manufacture rewrite them. Work on interlinking your internal pages like wikipedia does. Each product page should link to other product pages using good keywords in anchor text.
I recently came upon a similar issue as I noticed a link reported in Google Webmaster Tools pointing to my corporate website, from a site I didn't actually have a link on. But a form on that site uses a script hosted on my site, so it does look like Google sees those form action URLs and might even attach some value to them.
Seems logical to me in a way - in the crawling & indexation process the Google spider parses the full code and extracts all URLs.
I'm not sure if nofollowing them will help though. PageRank sculpting through nofollow has been hotly debated for a while now with compelling arguments on both sides of the fence. Personally I'm not a firm believer in PR sculpting with nofollow, but I love to be proven wrong. :)
Just because a link is reported in wmt does not mean it is being counted. They will also list mentions which is just your url in text with no link.
Thanks for the input, everyone ... every little bit helps, but I'm in agreement my time is better spent elsewhere rather than worrying about these.
"They will also list mentions which is just your url in text with no link. "
Yes because a search engine spider parses the whole HTML page and extracts everything it believes are URLs.
Would be good to test if non-linked plain text URLs pass any link value, or if spiders require an anchor for link juice.
I set up a cart hosted directly on a clients domain a few years ago and because I was concerned with duplicate content being in the cart I disallowed the cart in Robot's text. I'm no longer concerned with that content affecting the website so I recently un-disallowed the cart and the client's page count and links jumped drastically (understandably) and his #1 targeted keyword jumped from #4 to #1. So I do believe having a cart on your own site helps link count and page count and both of those affect your rank.