13 replies to this topic

[Pets]

I keep seeing that URL's for Dynamic Pages should be created to look like Static pages. I understand the good of this. Is there a negative aspect? Something which could hurt us as far as SEO goes?
Thanks

[robwatts]

I think you have to be careful and plan any changes.

For example, if you use mod_rewrite to change page.php?var=bar&var2=foo

to page-bar-foo.htm

You should ensure that requests for page.php?var=bar&var2=foo are handled correctly, in that a 301 RedirectPermanent points people to the new pages. This also ensures that your 'new' urls are not seen as duplicate content by search engine spiders and that your 'old' content is eventually dropped.

useful apache resource

asp iis url rewrite product

[Ron Carnell]

Let me prelude this by admitting that I've been working with computers long enough that I still count bytes when I'm programming, even though my servers all have more than the 8K our mainframes had when I started. In short, I'm very conservative (which sounds much nicer than stingy or cheap).

Having said that, I think mod_rewrite is a bad answer to this all-too-common problem. On a fairly busy site, the added stress on the CPU can quickly affect the load. On a fairly busy site, throwing in an aggressive spider to an already heavy load can bring a machine to its knees. My sites often tend to be seasonal to some extent, and there are typically a few weeks out of the year (leading up to Valentine's Day) when my machine might have to handle over a hundred times its normal load. I'm afraid mod_rewrite isn't even an option.

Fortunately, on an Apache machine, it isn't necessary. Suffice it to say, without getting too technical, that an URL of the form "/cgi/cpanel/action-display/page-author/" can be parsed in much the same way that an equivalent URL of "/cgi/cpanel?action=display&page=author" can.

However, in my opinion, that can be of limited value in terms of SEO and might even be detrimental.

We already know that Google, barring a Session ID, will index dynamic pages with at least a limited number of parameters being passed. Most agree that number is three. If you had more than three or four parameters, would it make sense to convert the query string into a static URL? Maybe. But my experience has been that the farther you move from your home page, the more difficult it is to entice Googlebot to index the page, and the more difficult it is to get the page ranked well. When every parameter in a query string is transformed into a sub-folder, you can get a LOOOONG way away from that home page very quickly.

I have nothing but anecdotal evidence to support this theory, but getting back to your original question, converting a dynamic URL with only a few parameters to a seemingly static one with a few sub-folders might well be counter-productive. I know Google will index and rank well a page in my /cgi with just a few parameters. Would I have equal success with the same page three or four folders deep? My experience suggests not.

[ResIpsa]

I convert using the ForceType method. Google has had no problem finding the pages and, from what I understand, “distance” from the home page has no bearing on PR. If hundreds of sites link to a particular page, regardless of the URL, you’ll have a decent rank.

[robwatts]

Hi Ron

I agree, the common perception is that Google does seem to handle upto 3 parameters pretty well.

I also hear what you are saying regarding getting down *too* deep in the directory structure.

Does anyone have any non-google observations on all of this?

I think one of the big draws of mod_rewrite was that around 18months or so ago, there was a belief that many spiders wouldn't touch, were scared of, or wouldnt read anyting that contained a query string. As things have moved on and capability has improved, then perhaps the need to clean up URLs has in fact diminished, or isn't at least as critical.

Interesting to hear what you say regarding server load, which I think is especially relevant on a shared platform. Although, that said, I think the issue may well be an ever diminishing one as hosts upgrade machines with bigger and better capabilities.

[robwatts]

Hi ResIpsa

Welcome to the forums.

If hundreds of sites link to a particular page, regardless of the URL, you’ll have a decent rank.

It is a pretty big *if* though. Very few internal pages get anywhere near that level of IBLs, in most cases they rely on the PR that is passed from the home page.

[ResIpsa]

I guess my point was that the "distance" from the home page doesn't (or shouldn't) matter to PR.

[Jill]

I guess my point was that the "distance" from the home page doesn't (or shouldn't) matter to PR.

I don't believe that the "physical" distance matters to PR. A page can be buried 50 directories deep and that's not a problem if that page is linked to from the home page. So yes, in that respect, distance doesn't matter.

Distance in the number of clicks it takes to reach a page, does matter to PageRank, however.

Jill

[ResIpsa]

But doesn't that only matter from a bot's perspective, not purely a PR perspective?

[Jill]

No, it does matter from a PageRank perspective.

If you look at PageRank (at least toolbar PR) on sites, you can see how it gets lower on the pages that you have to more clicks to get to.

Jill

[ResIpsa]

Isn't that just Google's estimate for PR? I know that the toolbar rank drops for each "directory", but I thought that was just a convention, not the real rank for the page.

Maybe I'm dead wrong, but why would Google penalize a page based on site structure, particularly when the page is not really buried in any directory? If these two pages have the same content and inbound links, they should rank the same, no?

http://www.mysite.com/script/12/10/4

http://www.mysite.com/script/12104

[Jill]

It's not a penalty at all. It's just the way PageRank works.

PageRank is all about the linking structure, that's all. The more links from higher pageranked pages any given page has, the more pagerank that page will have.

Jill

[ResIpsa]

That's my point. The URL doesn't matter.

[chill105]

This is really great stuff, and so helpful!

I was worried that google wouldn't crawl my page properly, but knowing that it usually handles three levels of variables is very good to hear.

Most of our pages don't go any deeper than that.

Thanks again, this site is great!

Zack