Do Deep Folders Stop Search Engines?
Folder depth comes up every now and again as a ranking factor (albeit a small one). The most recent mention is in a list of 22 considerations for improving natural search performance from Search Insider (written by Rob Garner of iCrossing). Some of the items he listed are accepted by the majority, but one in particular mentions folder depth as being an obstacle to search engines without additional support or detail. And so I'm taking it upon myself to see just how search engines handle folder depth.
Ideally I'd like to set up an experiment to measure the ranking benefits of a root-level page vs. one buried within sub-folders, but with so many variables to hold constant that wouldn't be easy. So to start, I thought I'd just focus on finding out if search engines will give up on files buried too deeply. My current take is that folder depth, within reason, won't stop a search engine from indexing content and that folder depth, again within reason, won't impact rankings all that much. I do believe that I once read that Yahoo cared about folder depth, but more current commentary suggests that major search engines are more concerned about click-depth i.e. the number of clicks away from the home page or some other authoritative page.
Some notes about the setup of this SEO experiment:
- I've created several pages that I've linked to from here. Having the links all on one page should eliminate some confounding variables.
- The folders are of varying levels with the deepest being 10. By the way, if you've got a site architecture that needs this many levels, you should really rethink things.
- The deepest folder structure is listed first. I did this because I know that a 3-level deep folder structure will get indexed so if this one does, but the others above it don't I'll be able to say that the search engines didn't index just the first couple of links in the list.
- I've removed the usual related links that appear at the bottom of my posts.
And here are the test links. There's really no point in you clicking on these, but I know some of you won't be able to resist.
[The Experiment Over – Links Removed]
I'll post updates about the indexing of the above URLs in Google, Yahoo, and MSN in a week or two, but I'll let things sit for longer than that before making any conclusions.
Eleven days after this post went live, Google and Yahoo had indexed and cached my test pages including the one 10 folders deep. Live hasn't cached any of them, but who cares about Live, right?
How do the levels rank? I am quite sure they will be indexed but just because they are indexed does not mean they will rank well?
It's true this test doesn't prove anything about ranking of folder-deep content. I suspect folder depth won't impact ranking much if the content is otherwise well-linked.
Thanks for the article! I recently had a site designed and am accustomed to all the pages at the root level. Instead the designer kept each page in an individual folder. I am wondering how this will affect SEO. I know that the file name is important but how important is the folder name? For example, domain.com/about/sem-consultant.html versus domain.com/sem-consultant.html. Will this have any effect on the optimization of the site. I am concerned with deep linking and having to change the name of the folder in the future. Why not have all the pages included in the root directory?
One level shouldn't make a difference either way.
Very interesting, would be great as well if you had entered some very individual (but identical to each page) words and then done a search positioning ranking comparision, they may be indexed but we could then compare folder depth effects in ranking.
Great post though, thanks.
The folders get indexed, which is great. Do meaningful folder names lend themselves to the optimization of searches? What impact, if any, do folders on sites have? If one of my urls was "www.eliz.com/great food/delightfulBites.com" versus "www.eliz.com/delightfulBites.com" would the url with the folder name come up higher on the list if someone were searching for good food? I know it's much more complex than this simple example. But your insight would be helpful in my understanding. Thanks.
Thanks for this, Marios. I was just dealing with a folder depth issue, and did a Google search to see if it was really an issue or not.
There's an interview with Matt Cutts that answers your question (so you don't have to take my word for it):
Eric Enge: Can a NoIndex page accumulate PageRank?
Matt Cutts: A NoIndex page can accumulate PageRank, because the links are still followed outwards from a NoIndex page.
Eric Enge: So, it can accumulate and pass PageRank.
Matt Cutts: Right, and it will still accumulate PageRank, but it won't be showing in our Index.
Here's the original article: http://www.stonetemple.com/articles/interview-matt-cutts.shtml
...what if for some reason I decided/needed to have a noindex on 1 of the folder levels - an example could be the a noindexed archive and the indexed single post the level below.
...does noindexed pages pass PR/linkjuice (or what ever you want to call it)?
I don't believe Google places a limit on the number of pages within a sub-folder it will crawl. Instead, I think Matt's article is saying that the number of sub-domains that appear in SERPs will be limited whereas in the past they weren't.
I appreciate your experiment; however, it doesn't address one of the points I read (with a degree of confusion, as it's a bit ambiguous) in Matt Cutt's (Google) blog. http://www.mattcutts.com/blog/subdomains-and-subdirectories/
I can't quite figure out from his post HOW MANY pages within a subdirectory/folder will be indexed--he refers to Google's prior approach versus "new"(written in Dec. 07) algorithm. There's an implication that Google may only index a limited number of pages WITHIN the folder. So I am interested in not only whether A PAGE your tenth-level folder is being indexed, but whether multiple/how many pages are being indexed?
Thanks for the info...
I have found that the lower level pages are indexed just as well.
It's a really good idea to do this test, because i spoked a lot with co-workers and no one give a prove if it's good or bad to the folders depth but now after your test i have a prove,
Thx for letting us know about this :)
Just wanted to let you both know that Google and Yahoo have indexed and cached the pages including the one 10 folders deep.
Excellent test. I look forward to the results.
I didn't bother with a root level page because we all know those get indexed.
As for the word count, that may effect rankings, but I don't see it having any effect on whether a page is included in a search engine's index (except possibly for really, really short pages).
Great experiment. I would've liked you to have included a root level page just to make a comparison between the rest of the folder depths. Also, is there a reason you varied the word count on the examples? Wouldn't have made more sense to make the word count as close as possible?