E-Commerce and "action=add&id123=" Links

[wmrobwl]

For the past 4 or 5 years we have been running a static (strictly html) web site with no funky symbols in the URL strings. We have enjoyed decent SERPS and rankings during this time. We have finally developed the e-commerce shopping cart of our dreams.

We are going to link the new custom built shopping cart to our current web site, leaving all of its current content, but just adding links to checkout and add to the cart. I am a little worried about implementing this because the URL string when you add something or checkout in the shopping cart contains "action=add&id123=".

I dont care if the SE's crawl the shopping cart, in fact I wouldnt even want them to. The problem is that I cant change the robots.txt page because there will be too many pages to list. Also the shopping cart resides on a different domain name. For example, if our web site was domain.com, our e-commerce website is secure-domain.com.

I am totally stumped, but there must be a way to stop the SE's from crawling this shopping cart.

Any help will be greatly appreciated. Thanks!

[brian.mark]

You could implement the meta robots tag to prevent your cart pages from being crawled. That should be easy enough.

Brian.

[wmrobwl]

So are you saying I should use this?:

<meta name="robots" content="noindex,nofollow">

And put it on every page that will have the shopping cart links?

I also read that currently only few robots support this tag. Is this true?

I also have the option of putting this code into the header of each page with products on it:

<script language="JavaScript">
function cart(action) {
window.open('https://secure-exoticracing.com/exotic/cart.lasso','cart','status=yes,menubar=no,scrollba rs=yes,resizable=yes,width=720,height=600');
return false;
}
</script>

If I use the java script, I understand that robots cant crawl java links, but what happens to the robot....does he just stop crawling the site when he hits the java, or does he just ignore it and continue crawling the site?

[brian.mark]

If it looks like a link, many bots are smart enough to follow them now. Googlebot is very good at finding links in JavaScript, especially if you're just using document.write, but also if you're opening windows. They've spidered a lot of our pop-up notices, and a few even rank rather well.

You could disallow *action=add* in robots.txt. Wildcard matches like that work with most bots.

Brian.

[wmrobwl]

Originally Posted by brian.mark

If it looks like a link, many bots are smart enough to follow them now. Googlebot is very good at finding links in JavaScript, especially if you're just using document.write, but also if you're opening windows. They've spidered a lot of our pop-up notices, and a few even rank rather well.

You could disallow *action=add* in robots.txt. Wildcard matches like that work with most bots.

Brian.

That sounds like a great way to handle it. One more question though. Can I disallow an entire site from being followed or indexed in robots.txt? For example:

If my site is domain.com, is there a way to disallow all links going to secure-domain.com?

[brian.mark]

If it is using the same set of web files, not really. If it's all SSL, you'll probably have httpdocs and httpsdocs (non SSL and SSL) or similar. You could put different robots.txt files in each directory.

If all the web files are grouped together in one directory, then unfortunately the answer is no.

Brian.

[wmrobwl]

Originally Posted by brian.mark

If it is using the same set of web files, not really. If it's all SSL, you'll probably have httpdocs and httpsdocs (non SSL and SSL) or similar. You could put different robots.txt files in each directory.

If all the web files are grouped together in one directory, then unfortunately the answer is no.

Brian.

The domain.com site does not have any links containing "action=add&id123=". The secure-domain.com site has the "action=add&id123=" links and the entire site is SSL (because its strictly used for the shopping cart.

I could make a robots.txt file for the secure-domain.com to disallow all bots, but then wouldnt I lose page rank on the domain.com by having so many outbound links pointing to secure-domain.com?

[brian.mark]

This is a terrible lie started by some uninformed SEO person many moons ago. You don't lose pagerank by pointing out of a domain to another one. It's page rank, not site rank.

Brian.

[wmrobwl]

Originally Posted by brian.mark

This is a terrible lie started by some uninformed SEO person many moons ago. You don't lose pagerank by pointing out of a domain to another one. It's page rank, not site rank.

Brian.

So should I just make a robots.txt page for the secure-domain.com to disallow *action=add*?

I'm also assuming since the domain.com doesnt contain any links with "action=add&id" in the URL strings, then disallowing *action=add* in the robots.txt of domain.com would be useless. Is that right?

This also means that since domain.com and secure-domain.com are two separate sites, then it doesnt even matter what I do, because I dont care if secure-domain.com gets indexed or not.

So then, I can also go the javascript route and can link each static product from domain.com to the shopping cart at secure-domain.com without having to worry about affecting any of the search engine crawlability of domain.com. Is this all true?

[brian.mark]

That's all true.

One word of caution from a usability standpoint, though: Not all users have JavaScript turned on. If you're selling stuff, JavaScript isn't the way to go.

Brian.