| 7:31 am on Aug 2, 2008 (gmt 0)|
It's a matter of judgement, and there's no clear consensus, however,
"This_is_probably_the_worst.html" possible way to do it - SEs do not parse "_" terribly well, and there's the added disadvantage that in an underlined link, "_" and " _ are indistinguishable, leading to write-in errors: neither machine nor human friendly!
"This-is-probably-the-best-way.html", as SEs happily parse "-" as 'space', it is clearly readable, and so has no disadvantage. Except that it is considered distinctly un-cuil.
See also: Quadrille's Oft-Quoted 14th Law:
More than one hyphen is international shorthand for idiot webmaster; More than two hyphens is Galaxy-wide shorthand for "I'd be a spammer if only I knew how"
"thisisstillconsideredcoolest.html", and SEs have learned to parse mergedphrasesreasonablywell. However, it is probably not parsed as well as "-" phrases, and few people tend to search for "thisisstillconsideredcoolest".
However, one school of thought says it really does not matter; this is one very minor point out of 200, and is dwarfed by such key items as, er, keywords in text and anchor text.
The Tesco school of thought says "Every Bit Helps".
Either way, I am 99.9% sure that whatever you do, do it for new pages only; any benefit would never compensate for the dame caused by changing URLs.
| 9:39 am on Aug 2, 2008 (gmt 0)|
Many thanks for your prompt response on this point.
I totally agree with you deleting a page once spidered key words and too many "-"'s are very important.
The aim is to leave old pages in place and implement a site wide policy for future naming of pages and perhaps duplicate existing ones with the new convention when the page content is next updated.
If I can offer assistance for your shared wisdom please shout
| 2:08 pm on Aug 2, 2008 (gmt 0)|
thanks for that info I found it very interesting as a newbie to this forum.
| 4:40 pm on Aug 2, 2008 (gmt 0)|
DVD-Replication.htm is friendly with Google Bot.
| 9:05 am on Aug 3, 2008 (gmt 0)|
thanks bilalseo, how do you test to know if something is friendly, does google have a page as it has so many webmaster tools.
| 1:58 pm on Aug 6, 2008 (gmt 0)|
|More than one hyphen is international shorthand for idiot webmaster; More than two hyphens is Galaxy-wide shorthand for "I'd be a spammer if only I knew how" |
I am curious why you would have this opinion. Earlier in your post you mentioned...
|"This-is-probably-the-best-way.html", as SEs happily parse "-" as 'space', it is clearly readable, and so has no disadvantage. Except that it is considered distinctly un-cuil. |
but then you say this is poor design style. I was wondering why this is considered a bad practice.
| 2:19 pm on Aug 6, 2008 (gmt 0)|
The second quote is about SE ability to parse, and what is therefore (marginally) preferable to use for maximum SEO benefit.
The first quote (a little tongue_in_cheek), relates to the *appearance* of the URL.
Sadly, while most honest SEOs have been loyal to stringemtogether, many sites lower down the food chain have welcomed-the-hyphen with open arms.
So, somewhat unfairly, the hyphen has become inextricably linked with get-rich-quick schemes, overpriced e-books and other less reputable sites, further damaging its attrcativeness to quality site builders.
So while I believe the hyphen to be the most SE-friendly of 'space proxies', I fear that overuse and the impression thereby created would negate much of the benefit - if not all.
Hyphens, like .info, have a reputation. And reputations damage all, not just those who deserve it.
I do use hyphens. But I consider possible consequences first.
[edited by: Quadrille at 2:20 pm (utc) on Aug. 6, 2008]
| 3:21 am on Aug 7, 2008 (gmt 0)|
|So, somewhat unfairly, the hyphen has become inextricably linked with get-rich-quick schemes, overpriced e-books and other less reputable sites, further damaging its attractiveness to quality site builders. |
Ah, I see your point. Yes, unfortunately I have seen the same practice before but never associated it with the hyphens, but I see exactly what you are saying. So a better question is all things considered what is the best way to do it [name pages and be SEO friendly] if you really do need to add a couple of words and have them parsed well.
I have a few sites where the amount of pages are so large that I really need some type of naming convention that will help me sort out the pages for future editing and of course I would like to SEO friendly if possible.
| 8:55 am on Aug 7, 2008 (gmt 0)|
The 'seo benefit' of putting key words in URLs is probably very small; it's one factor in many, and for some reason it is attracting waaay more attention than it deserves.
Yes, 'neat' urls do well in the serps. They also do badly, or in between, depending on much more important factors on and off the page.
And long URLs eitheralongsillyphraseruntogether, or a mixture-of-random-words-linked-by-hyphens is very far from human-friendly.
I try to combine the best of all worlds;
This allows me to use the brand (domain name), the 'local' key terms (file name), and one or two relevant key words (folder names).
It also helps me to keep site navigation logical and user friendly.
Most of all, it means I don't obsess over it; I have a system that requires little thought.
probably just as well, in my case ;)
| 6:44 pm on Aug 8, 2008 (gmt 0)|
|This allows me to use the brand (domain name), the 'local' key terms (file name), and one or two relevant key words (folder names). |
Great suggestion! Thanks!
| 7:11 pm on Aug 8, 2008 (gmt 0)|
Whatever, make sure you avoid underscores and spaces.
Additionally, mixed-case URLs can also be a problem.
I prefer using dots between words.
| 2:57 am on Aug 11, 2008 (gmt 0)|
Great info, thank you all for this useful thread.
| 1:22 am on Aug 12, 2008 (gmt 0)|
I have to agree with Quadrille. It's really not as important as most people make it out to be. It has much less affect on your ranking than other pieces of the pie. But if you're trying to squeeze the last bit of blood out of the turnip, selecting the proper URL naming conventions can help a bit.
That being said, I am in the process of converting our large commercial site that has about an 80+% brand awareness from an .ASP/.ASPX platform to a CMS where the the names of the pages will no longer have a file extension. So essentially every page (except the home page) will eventually have it's name changed. We have the new site up in Beta at the moment...
Since every page on our site (25,000+) is going to have it's name changed anyway (dropping the extensions), I took this opportunity to set some standards for URLs.
1) All lowercase URLs
2) Use keyword rich URLs - usually 2, maybe 3 words... (our new CMS only allows a 32 character name for a page) but made exceptions for articles since we have so many and they are built specifically for attracting traffic from 4,5, and 6 word long tail searches... They were previously named with the title of the article which sometimes resulted in titles like this-is-my-article-about-something-silly-that-brings-lots-of-long-tail-traffic.aspx for example. Many articles with 10 word titles separated w/ hyphens rank fine and generate a lot of traffic from Google and other engines. It looks tacky, spammy... but it was done innocently as previously a business owner would actually determine the page name in a cheesie article database we used to build static pages.
3) Use hyphen as word separator - no underscores, no cramthewordstogether, no spaces...
4) All branches in our site tree are followed by '/' without exception. Only leafy nodes in the site tree (pages that will never have pages under them in the site planner) are not followed by '/'.
Since we are getting away from .ASP/.ASPX I didn't really have to worry about default pages for a folder anymore. But if we were using .HTML, .ASP, or .ASPX I would have made sure no links on our site explicitly specified index.html or default.asp or default.aspx in the URL. All default pages for a folder would have been referenced by referencing the folder followed by a '/'. For example, to run http://www.example.com/folder1/default.asp I would have referenced it as http://www.example.com/folder/ everywhere. It doesn't matter whether you include the default page name or always end such URLs w/ a '/'. Just be consistant all over the site.
| 3:19 am on Aug 12, 2008 (gmt 0)|
It's a Work of Art!
May I quote you? (with full attribution, of course!)
You need to add, for the non-techs (including me), that there's ways to make that change without having to start again with the SEs.
[added] I absolutely agree that 'every bit helps' - but I do worry when one of the minor points get raised to saintly status[/added]
[edited by: Quadrille at 3:22 am (utc) on Aug. 12, 2008]
| 5:19 am on Aug 12, 2008 (gmt 0)|
I agree with dumping "file extensions" from page URLs, but I'll stick with the HTTP standards:
And being a bit more pragmatic, why make your type-in users add the extra trailing slash?
For those new to the subject, using extensionless URLs for your pages requires a bit of behind-the-scenes coding on your server. There are several ways to do it, but in general, some mechanism is needed to associate the extensionless URL with the proper file -- which generally must have an extension, since it is a file. This can be done with Apache mod_rewrite, ISAPI rewrite on IIS, MultiViews/Content Negotiation, Apache AcceptPathInfo, cgi scripting, PHP, or a combination of these or similar server features.
A key point in thinking about this is that URLs are not filenames and filenames are not URLs; These are two different naming conventions, one for use on the Web, and the other for use within a server. A bit of server code can easily tell the server, "When you get a request for a URL with no "file extension," then add .html (or .php or .aspx as applicable) and go serve that file."
| 1:12 pm on Aug 12, 2008 (gmt 0)|
Our new CMS system automatically renders branches (the equivalent of folders) w/ a trailing slash and leafy nodes (the equivalent of an end node page) without a trailing slash. I'm digging it. 1st time I've worked with a CMS. The business owners are loving it as well. No more waiting on a developer to make a simple text change or to add a widget to a page.
The real challenge has been coming up with a way of changing 25,000 URLs without totally disappearing off the search engine map. We have a very large section of articles on the site that we plan to use URL rewriting to display the new page with the old URL for some time. Only a small number of core pages will be 301'd initially. Then at our leisure, we will convert the URL rewrites to 301s section by section (adding [R=301,L] to the rewrites) so that at any one time only a small section is taking a hit while the engines recrawl all of the sites linking to it.
Scary as hell... but should be fun if it works! :)
| 2:03 pm on Aug 12, 2008 (gmt 0)|
I'm a great believer in the 'gradualist approach' - I always reckoned the Borg could assimilate small groups much easier than whole planets - and what's good enough for the Borg is good enough for Google :)
| 8:57 pm on Aug 25, 2008 (gmt 0)|
enter in google search:
and you will see results and all will be clear ;)