In this report we need only be concerned with Generic and Country Code TLD's for SEO purposes.

Everything in the URL beyond the TLD —i.e., subdirectory, filename, and extensions— are all case-sensitive. In other words, index.html is not the same as Index.Html

It's common knowledge that the root of SEO is keywords. Since the beginning, all strategies have revolved around the ever-changing and constantly evolving targets for keyword placement. In the swirling sea of anchor links, meta tags, headlines, body copy, and what-not, one of the most consistently useful placements in terms of SEO has been within the URL itself. And the degree to which this strategy remains effective is simply a matter of how far the engines scale-it-up on the algorithmic dial.

This is why the use of keyword-specific subdomains has been a long standing and typically effective SEO strategy. When viewed from the search engine's point of view (SEPOV), this also makes sense. That's because when a keyword is found within the URL there is typically a very high probability that the keyword is relevant to the webpage's content—and search engines hunger to provide relevant results.

So, keyword placement in the subdomain accomplishes two things:

1. It provides a clue to a potential site visitor what the site is about prior to clicking the link in the search results (as would be the case with a domain like CheapAirlineTickets.mysite.com).

2. It gives the search engines an additional relevancy indicator which they can choose to use—a little, a lot, or not at all—within their overall relevancy algorithm. Over the years, we have found that search engines cannot resist using it to varying degrees—even if only a little.

Think about it like this: If all else is equal (which is only hypothetically possible), the site with the keyword in the subdomain will likely rank higher than an otherwise equally optimized site without it. And, perhaps more importantly, given a choice between two otherwise equally attractive selections within the search results, the average potential site visitor will choose the link containing the keyword in the subdomain over the link that does not.

Utilizing a subdomain is one of the best legitimate strategies for placing your primary keyword into the URL when the specific keyword-TLD name is already taken or otherwise unavailable.

Currently, Google, Yahoo, and MSN each appear to be giving some ranking boost to pages that contain the keywords in the subdomain of the URL. We believe that Yahoo places the most significance on keywords in the domain name, closely followed by Google. MSN doesn't seem to place quite so much emphasis on this, but it does appear to factor it into their scoring.

We can't help but notice that most of the top pages in the search results for very competitive searches contain the keyword either in the domain or the subdomain name. One good example is a search for music. On each of the big three engines you'll find many of the top results have the keyword music either in the subdomain, primary domain, or in some cases, the subdirectory—or else you'll find a keyword (like mp3 or mtv or itunes) that is synonymous with music.

Although there isn't any question that keywords in the subdomain can help rankings, some cautions are in order. Taking into account that subdomains can be separated by a period—making possible certain combinations like keyword.keyword.domain.com—we recommend sticking to the more customary single level format; keyword.domain.com. Rarely do you see multilevel subdomains ranking for the more difficult generic searches.

Another possibility is to delimit your subdomain using dashes. For example keyword-keyword.domain.com is also a technically available possibility. However we don't recommend using more than one dash in a subdomain even though we have seen some good results at MSN using more than that. Regardless, unless you're optimizing solely for MSN (unlikely) you should limit subdomain dashes to one, or none.

A keyword placed in a subdomain is not only a ranking factor, but also a linking factor. Frequently, when another site links to your site, they use the domain name as the anchor text of the link. Placing your best keyword in the subdomain means that other sites will be compelled to use that keyword in their link to you. That will score your page some extra relevancy points from the search engines point of view.

Another caution to be aware of is wildcard subdomains where anything.domain.com results in the same page as anythingelse.domain.com. Search engines have a VERY dim view of this practice and we don't advise using it.

Here's another warning: Don't create a subdomain when there isn't a good user focused reason to do so. Avoid having subdomains with only one or two pages on them—a small number of pages on a subdomain (other than www.) is a red flag to a search engine. You can expect that ranking penalties or outright bans could be levied on any site that combines the typical www.domain.com format with a bunch of low-populated keyword-laced subdomains in an obvious effort to manipulate rankings.

If in doubt, it's always a good idea to review how some of the major search engines are structuring their file systems. For example, look at Google. Most of their site is at www.google.com. However they assign a subdomain to certain specific and large areas of their site. For example news.google.com, groups.google.com, froogle.google.com, and local.google.com.

Each of these subdomains are logical separations and, clearly, they have good reason for dividing their site into these subdomains. A contrasting example would be a shoe site that uses nike.sitename.com, adidas.sitename.com and so forth. We're not saying this wouldn't be effective, only that it's pushing the envelope and the next algorithm tweak might land the whole site in the penalty box. If you do it, be sure to place plenty of unique content within each of the subdomains. Otherwise it's sure to be viewed as a spam technique—so beware.

Your domain, of course, is your registered domain name. For example, this site's domain name is SearchEngineNews. Naturally you don't have the flexibility of modifying it in the way that you can with subdomains. Regardless, the ranking advantages enjoyed by keyword-smart subdomains also apply to domains.

Clearly, it helps considerably to have good keywords in your primary domain—both from a search engine and a consumer perspective. The ranking boost is most profound when the domain name exactly matches the keywords being searched on. However, due to past abuses, there are some unwritten restrictions and warnings to be aware of.

At one time, keyword-keyword-keyword-keyword.com had a boosting effect on rankings, but that didn't last long. The engines were quick to realized they were being gamed and began counting hyphens and domain name character lengths. As you might expect, they were able to correlate multiple dashes and long domain names with spam sites. It didn't take them long to restrict how high a long, multi-dashed domain name could rank.

Today we recommend no more than one dash, and the shorter your domain name, the better. In a perfect scenario, your best domain name is typically your primary keyword or keyword combination. Whenever that isn't possible, at least try to get the most important keyword mentioned somewhere in your domain name. Otherwise, expect to settle for placing it into your subdomain.

To hyphen or not? A few years back, hyphenated keywords within some domain names enjoyed a small ranking advantage. Hyphens were used to delimit the text so the search engines could more easily distinguish each keyword within a phrase without mistaking it for some unique term. However, today we're seeing more indications that the major engines are getting better at picking keywords out of a phrase without the help of dashes—at least in the English language.

So today, whenever faced with a choice, you should favor the keywordkeyword.com over the keyword-keyword.com URL. While it's true you may want to secure both URLs to keep the other one out of the hands of your competitor, you would be better advised to develop the URL without the hyphen as your primary site.

If you can't get the keyword into the domain name and it isn't advisable to put it into the subdomain, then your next best option is to place it in the subdirectory name.

Bear in mind that it really doesn't matter if you use a subdirectory or a file name to contain the keyword, but typically one (but not both) will help search ranking. In most cases, /keyword/index.html, ranks equally with /keyword.html.

Regardless, you will likely have better results if part of the file structure contains at least some part of your keyword phrase. Subdirectory names are also sometimes important for other reasons. For example, take /cgi-bin/. Historically that's been a subdirectory avoided by search engine robots for fear of getting trapped in an infinite loop and indexing millions of unique URLs that are actually just duplicate pages.

For the most part, however, search spiders have solved this problem and will index URLs that contain cgi-bin. Still, we suspect some limitations still remain, so, ideally, you should avoid using that specific subdirectory name whenever possible.

Other similar directory names that may have built-in limitations are popular software program names like /phpBB/ for the PHP Bulletin Board software and /Gallery2/. If you install one of these programs we would recommend that you use a unique directory name whenever possible just to ensure you'll avoid whatever limitations to indexing that might still be lingering.

Just as with subdirectories, having the keyword as the file name is usually a good idea. However, be forewarned that you don't want to overdo any part of this. One mention of the keyword in the URL is often good enough. Multiple repetitions in the URL are typically associated with spam and ranked lower.

For example http://www.domain.com/keyword.html is ok, but http://www.keyword.com/samekeyword.html is probably overkill. Examine the search results for your keywords and look at what is scoring at or near the top. You'll see the pattern quite clearly as to what is good and what isn't.

This is a frequent question that pops up in our tech Q&A's — Does the file extension affect my search ranking? Typically NO, it does not, at least as long as the extension is one that is commonly associated with a web page.

We have not seen a case where this mattered at all in the last couple years. Of course, .html or .htm is the most often used file extension. But more and more we are seeing file extensions in the top 10 search results that include .cfm, .php, .asp, and .aspx. Ranking-wise, all of these file extensions are equal in the eyes of the engines.

In general, URL structure for SEO optimization follows a general rule that...

the more generic your keyword, the earlier you want it in your URL structure.

For instance, if you want to score for the extremely generic term music that returns more than a billion search results at Google, you should definitely place it in either your domain or subdomain name.

However, if you want to score for a specific model number for a cell phone your keyword wouldn't be anywhere near as popular as the keyword music. So, using your keyword—the model number—as a subdirectory or file name will typically work quite well.

Also, you should be wary of over-doing it. A good rule of thumb is this:

If the URL looks like spam, it probably will be treated as spam.

The search engines caught on a long time ago to the www.viagra-pills-porn-casino.com/ style domain names and such similar file structures. Today you want to use domain and file name structures that appear to be common sense to the human visitors of your site. Always bear in mind that people do look at the file names within the search results. And seeing the keywords highlighted in the URL does increase click-through-rates.

A static URL looks like this:

A Dynamic URL is typically characterized by having certain characters like ? after the file extension. For example:

Everything after the ? mark is typically a variable.

Remember, many dynamic systems do not need all of their variable information in order to work correctly. Many programmers go overboard on the variables they add to the URL structure.

Today's major search engines have added capabilities for crawling dynamic URLs that weren't available a few years ago. Still, there are limitations. A site rarely gets fully indexed when there are more than three variables in a dynamic URL. Ideally you should avoid using more than two.

There are also some specific variable names– like ID – that indexing-bots often avoid. That's due to what are called session variables—a variable that is unique to each visitor to the site. The use of session variables often results in many, many, duplicate URLs being indexed in the search engines. That's why the bots do their best to avoid URLs that appear to have a session variable.

If you are operating a dynamic site you should, ideally, avoid using session variables whenever there is an alternative. And, if not, find a way (like using robots.txt) to prevent the search engines from crawling those URL paths.

The most effective way to prevent indexing problems with session variables is to use a good IP delivery program that will recognize search engine spiders and make sure they only receive URLs which have had the session variables removed.

Note: a large number of URLs with session variable have begun appearing in Google recently. This appears to be a glitch in Google's indexing process that should be resolved soon, as Google states on their webmaster guidelines page that they do not index pages with session variables.

It's actually a major headache for many webmasters, since it means that many web pages are getting indexed multiple times under different URLs. All the more reason to employ a solution which prevents search engines from being served session variables.

Another question that often comes up about URLs is how to refer to them in the HREF section of a link. There are two options here: absolute or relative. An absolute link means you use the entire URL:

http://www.domain.com/filename.html

...where a relative link simply refers to:

filename.html

While either link will work just fine when referring to pages on the same domain, absolute URLs are preferred. They're better for a few minor reasons, like...

• The links will work if someone steals your content or saves it to their desktop.

• Absolute URLs help avoid getting a dumb-bot stuck in a loop that increases server load and generates 404 errors.

Admittedly, it's a minor point. But, if you're striving toward SEO perfection, absolute links are part of the perfect package. Anything you can do to make it easier for search engines to index your pages is a good thing.

Keywords in URLs can help rankings, but remember that changing your existing URLs without redirecting it to a new one will break all the links and bookmarks pointing towards that URL, causing the page to drop out of the search engines.

This means that if you have an existing high-ranking page, it's almost never a good idea to change the URL. If you must change it, be sure to use a 301 redirect to make sure links, user traffic, and search engines are properly sent to the new URL. For a complete tutorial on using the 301 redirect to change URLs, see our report:

Unraveling the Versatile 301 Redirect