Directory Stucture

Forum Moderators: open

Message Too Old, No Replies

Directory Stucture

How much damage is this doing

hurlimann

6:34 pm on Nov 10, 2002 (gmt 0)

I keep telling my coders to do pages with the content near the root. They keep doing this:

http://www.anything.com/sub1/sub2/product.html

Sometimes sub1 and 2 are parts of keyword phrases we use.

Is this worth demanding the change and if so what to?

WebManager

6:37 pm on Nov 10, 2002 (gmt 0)

Keep it natural and it will be o.k.

i.e. does it help the surfer and is it a logical structure.

andreasfriedrich

6:38 pm on Nov 10, 2002 (gmt 0)

PR is not based on your directory structure. So you should be ok.

Andreas

Nick_W

6:41 pm on Nov 10, 2002 (gmt 0)

Yes, not related. It's your links that count.

Naming is great though. I've had lots of luck with:

/fuzzy-blue/widgets.html

and similar....

Nick

Terry_Plank

9:47 pm on Nov 10, 2002 (gmt 0)

We advise that down two levels is the better then lower. Now, it is not going to be true in every case so there will be examples of lower and higer getting ranked because there are many different variables. But, I still think that as a general rule I would still try for no lower than [anything.com...]

europeforvisitors

9:58 pm on Nov 10, 2002 (gmt 0)

We advise that down two levels is the better then lower.

Does Google even care about directories? Googlebot isn't drilling down through a site's directories; it's following links. So it shouldn't make any difference is a URL is:

[bakedgoods.com...]

Of course, the above example assumes that both URLs are linked from the same page (.g., the home page, a site map, or a table of contents).

Side note: I've got pages in subdirectories that perform better in Google than other pages that are at the root level, probably because they're linked from my navigation bar or from the home page.

Marcia

12:19 am on Nov 11, 2002 (gmt 0)

hurlimann, is it primarily Page Rank you're concerned with?

>>Sometimes sub1 and 2 are parts of keyword phrases we use.

Aside from PR considerations, there's some argument in favor of having keywords in URLs, in the file path, that it helps with rankings. Some believe that having keywords in the domain name helps with rankings, whether it's giving a boost in itself or because of inbound link text. In cases where the domain name is for branding rather than having keywords in it, using subdirectories gives the opportunity to have the keywords in link text, primarily with the site's internal navigation, but also for external links at times.

Staying close to efv's food example, let's say it's a food site with dozens of categories and you have

[foodexample.com...]
[foodexample.com...]
[foodexample.com...]
[foodexample.com...]

You can have a link to the meat section from the index page, which puts the keyword in link text and URL, and then there will also be link text for the sub-categories from the meat section. This seems a little harder to do when all the pages are in the root directory.

From a PR standpoint, if they're all in the root, it might not be wise to have them all linked to from the index page, which can be a waste in terms of PR distribution. If there are 100 people a day looking for beef and 10 a week looking for veal, and beef is much more competitive than veal, beef will most likely need more PR to compete and be worthy of more attention. Veal can be linked from the /meat/ directory, but if beef is a high value/traffic item, it can be linked to from the index page as a featured item in addition to the link to the meat category.

By doing that the keyword that needs higher PR is then one click away from the homepage, which most likely would have the highest PR, and the less valuable keyword is 2 clicks away. It's not the directory structure, but the number of clicks away from high PR pages that maximizes or diminishes the PR distribution in a site. This may be hoarding PR in terms of internal navigation, but it's also closer to meeting user demand.

>>>pages in subdirectories that perform better in Google than other pages that are at the root level, probably because they're linked from my navigation bar or from the home page.<<<

On one site I work with there were a lot of pages for new products added to the root directory, all of them linked from the homepage and main navigation, some of which result in no benefit for the site in terms of revenue. They caused a drop in PR to very important pages in terms of relative consumer demand and profitability, which are in /directories/ and had PR equal to the homepage. It didn't affect rankings, but the less valuable pages will be taken out of the root and put into a /directory/ with a good keyword that'll be linked to from the homepage and main navigation and those lesser pages will be linked to only from that directory and one or two closely related pages topically, NOT the homepage. They don't warrant anything better than 2 clicks, whereas the keyword phrase for the new directory does - it'll be promoting for an additional product line in a way that's not diluting the PR. It'll keep the number of links off the index page to other site pages or section minimized for maximum advantage.

Use of directories can be very convenient in certain respects, for PR and also for meeting user demand with logical navigation. This is where watching traffic and sales figures over time can help; you can actually take that data and sit down and flowchart the site navigation for PR distribution to the most desirable pages.