Pages in Folders and # of Clicks from home page

[St0n3y]

Several years ago it was common SEO thought that you wan to keep your pages out of sub-folders and as close to the root directory as possible. I have not heard or read anythign about that in some time and I'm wondering if that still is the convential wisdom or not.

I would like to get some opinions on what would be more search engine friendly:

a) Page X is located in the root directory but can only be found 10 clicks away from the home page.

b) Page X is two clicks away from the home page but the file is located in site.com/folder/folder/folder/folder/folder/page x

My gut is to think that its the clicks that matter, not the folders. Any agreement or disagreement with that?

[greenleaves]

My gut agrees with you that it would be the clicks and not the folders that would count.

When it comes to putting pages in folders, I think this is something that should be done to organize your site properly, and that SEs are not what you should be thinking of when creating the folder structure of your site.

[telNform]

IMHO - I believe that files within the root of the domain will have a higher weight associated with them (all other factors being equal - of' course). I am not saying there is anything wrong with organizing your files within directories, but I would not allow empty directories within your site architecture either.

[St0n3y]

another thought that occurred to me, assuming Google will only download so much information about a site, would the extra folders compromise the amount of info that will get downloaded?

[seomike]

Just look at the pagerank loss in directories with deep folders Yahoo being a perfect example.

After building a directory that mirrored the exact structure of Yahoo's directory I found a couple of things out the hard way.

1. I couldn't get a spider to touch the deep folders.

2. I couldn't get spiders to crawl 3 folders deep until I got a PR of 3 or higher on the home page.

3. PR typically drops off a point as you go deeper. root = 5, sub1 = 4, sub2 = 3 sub4 = 2 etc.

4. spiders will not crawl 10 folders deep unless you have a huge PR page linking to it. And i'm talking a page that has like a PR of 8 or higher.


I keep my directories down to 1 sub folder 2 at the most and I have had much success. :D

[St0n3y]

That sounds about right, but you do you think that might have more to do with the link structure rather than the folders? If those pages many folders deep were still accessible just a couple of clicks from the home page, do you suppose you would still see the same thing?

[Robert_Charlton]

My gut is to think that its the clicks that matter, not the folders. Any agreement or disagreement with that?

I believe it is the clicks, not the directories. If you take a page in a deep directory and link directly from the home page (or link to it globally), it will most definitely get a boost.

That said, if you put too many links to inner pages on your home page, you're losing the ability to intelligently control your PR distribution.

Use your pages closer to the root to target the more general, more competitive phrases... and, as you go deeper down, target the longer, more specific, and less competitive phrases.

Also, take a look, eg, at the dmoz home page, and note how they prioritize. They link not only to the main categories, but also to some of the most important subcategories from their main menu. Eg, under the link to the Science category they include links to several of the most important specific science categories... Biology, Psychology, and Physics.

I'd avoid too deep a directory structure for other reasons... and perhaps there may be spidering issues too... but in my opinion it's your link hierarchy that matters.

[Anthony Parsons]

From my experimentation, it has nothing to do with folders deep at all, and more to do with clicks away. Google finds just about everything, but Yahoo is another story. If you have a standard folder depth for example:


.com/folder/folder/folder/folder/folder/

If each folder is only linked from the one above it, then yes, the spiders get weary to go to much deeper. Whether that be they are programmed to avoid spider traps or place some sort of relevancy upon information buried that deep, don't know.

In my experiments in the past, Google will follow the above without an issue, Yahoo on the other hand, basically what Mike said, however; where I have had more success is by ensuring that everything is listed on a sitemap making it all only two clicks away from the homepage.

For example, if you look at UCN directory, the sitemap is an alphabetical listing of all categories / sub-categories in which each letter of the alphabet is listed on the homepage, which means each spider can travel the normal route to get to the page or simply hit a letter of the alphabet, one click to the alphabetical listing, then the second to each cat / sub-cat starting with that letter. It works, and it works a charm.

Sounds silly, but directories need sitemaps to ensure the spiders do go deep and the directory is managable without everything stuck as a root page. Not that this is an issue so much for most directories being dynamically generated, but we must make it look the part for the spiders.

[Dave Hawley]

I think the saying "do what's best for your site visitors" would be apt in this case. If that means deeeep folders, then I guess a site-map could be used to help spiders out.

[seomike]

That sounds about right, but you do you think that might have more to do with the link structure rather than the folders? If those pages many folders deep were still accessible just a couple of clicks from the home page, do you suppose you would still see the same thing?

If you were to link to every page like this from your home page you would bleed the home pages PR and therefore drop the PR passed to the deeper folders.

This is why links from a site map page pass very little PR.

The only real way to counter PR bleed in deep directory structures is to get external links into the deep folders. This way you have PR flowing up and down.

[St0n3y]

I've seen sites that are built with every page in its own sub-folder. One site in particular had dynamically generated pages, but linking to the folders rather than to the "pages" made for much more friendly URLs. Valid strategy in such cases?

[Robert_Charlton]

I think the saying "do what's best for your site visitors" would be apt in this case. If that means deeeep folders, then I guess a site-map could be used to help spiders out.

Good point, Dave. My slant on it, one which I feel should be emphasized, is that SEO, done well, optimizes for users as well as for search engines.

Part of what search engines do is try to translate page and structural relevance into ranking relevance... to translate human perception into "machine" terms. Thus (apart from anti-spam considerations), most SE algos attempt to give more weight to what is visible and prominent on a page or in your site structure.

If you know what's most important to users, you should make that more prominent on your site. In your linking structure, eg, your most important links are liable to occur high up in your link hierarchy, to be most visible, and to be global within relevant areas of your site.

Thus, my dmoz example...
They link (from the home page) not only to the main categories, but also to some of the most important subcategories from their main menu. Eg, under the link to the Science category they include links to several of the most important specific science categories... Biology, Psychology, and Physics.

Assuming that your links are links that spiders can follow, a navigation system that prioritizes well for your users should also serve best for spiders.