I have always wondered, if the file name, which becomes part of the url itself, mattered as it didn't seem to, and no one had tested the matter ... until now ... Apparently under the right conditions, it does count for keywords if the url string is set up right, which starts with what you name and save your html file as.
At one time I believed as did/do most of it us, that the file name of a page was nothing more than personal convenience for keeping track of " what is what " in your own files, so whatever worked, was fine. We have all seen url strings that look like they were written in binary, that work just fine.
However, according to testing, if you want a small boost of keywords for that page, you can and should use the file name for keywords, as much as you do the title, the text content or keywords in your meta tags.
Ok how ?
First off lets look at what is commonly done
Which is my " all as one word string " for the lead page for these articles. Useful for me, as its fairly easy for me to know which page I am looking at, in my own files, even if all strung together like that can get a bit cumbersome to read ... its not of any value for the engine's for keywords however. This is one of the most common you will see.
That string, as individual words, is also in the pages title and meta tags, so I'm not losing any value, just not getting any extra value by having the url itself, set up differently.
A space between the words, with no hyphen or underscore mark ... which is how I used to do it, makes it easy to read as a file name, but makes it bad as a url , as it fills in the blanks for itself.
That %20 is url speak for a blank space ... looks weird doesn't it.. :) but the Web can't handle blank spaces, so it has to fill it in with something, and does so automatically.
Now on my personal files, just looking at file names on my own hard drive, it looks like this " Fibromygia the unbelieved ", which is a lot easier to read than Fibromygiatheunbelived . These spaces work just fine for an internal link, and it's recommended to do just that for " anchor " text.
I started doing the run it all together however, to avoid that %20 bit in all my urls, once I realized why it kept doing that, ... but, while that helps to avoid this issue, it doesn't help for keywords for the page and it's clumsy to read.
Underscore _ which is a common variant for a spacer, often seen in email addresses for example , apparently forces the engines to see whatever is on both sides of the underscore as one word.. so not of any value again.. even if its easier for us to read.
So how to use key wording in your file names, to where it means something to the engines according to all the tests ?
Which is the first file name, I ever did that with, just as my own personal test case, and have continued the practice with every new page I have built since, that has a file name that is over one word long.
Why ? Because it does work. I purposely put that page out there to start with, with no title and no keywords in the meta tags at all, letting the url alone carry most of the load for keywords.
And then went to find it. No problem, for any of its keywords and combinations. I have since put in the proper titles and metas for it, but wanted to see, on a page that I built solely as a temporary addition, how well it would work. I am happy to say it works just fine.
A plain common hyphen mark - These keywords were picked up, by the engines, as individual words right off the url itself, as the hyphen is the only acceptable separator in a url, that the engines recognize. And its easy enough to read as a file name.
So for new pages, it's worth considering doing for an added boost of keywords based on the url name string ... so something to think about when your naming your newest web page's file name.