SEO and file names

[ImagineThat]

I am curious to get additional input on the imortance of file names (i.e. keyword.html) as it relates to SEO. Some experts advise to name website files using keywords, and others are silent on the subject. Given the formula for page density, # of inbound links and the actual text of the inbound links, I don't know where the name of the file can fit into the equation. Thanks in advance.
Teri

[pleeker]

I am curious to get additional input on the imortance of file names (i.e. keyword.html) as it relates to SEO. Some experts advise to name website files using keywords, and others are silent on the subject. Given the formula for page density, # of inbound links and the actual text of the inbound links, I don't know where the name of the file can fit into the equation. Thanks in advance.
Teri

I've only recently been testing this on a couple client sites -- using file names such as electronic-blue-widgets.html and such. From what I can see from this limited experiment, it can help, but I'd never suggest it's one of the key elements of a web page (such as the <title>, meta description, H1s, etc.). I would think that if the keywords in the file name can also be found in the <title>, the Meta Description, the H1s, and the main body text, there could be a slight boost to a page's relevancy for those keywords.

I would love see some responses from others who have better knowledge and evidence.

[Nick W]

I've found that paths are pretty cool in some circumstances lie if you sold blue fuzzy widgets:

/widgets/blue/fuzzy.html

Im not an SEO really, but I pick up a thing or 2 here and there and I know what works for me, and that certainly does work for me ;-)

Theme pyramid?

Nick

[rustybrick]

Here is the most recent study that I am aware of from Alan Webb's Forum (http://www.abakus-internet-marketing.de/en/forum/viewtopic/t-775.html)

I've decided to open some research for public scrutiny
It involves the domain webmaster-community.com. targetted keywords are "webmaster community" which started off ranked 19th.
The ranking report before I made before any changes based on one page frameset that loaded my german forum can be found here...
http://www.abakus-internet-marketing.de/rreports/seotest.htm

The "optimized" page just got indexed and there is a change in the ranking which I will report in half hour or so.

I have one big request, please do NOT link to the page (yet). I want to test onpage optimization first and then build links to it. Currently only 4 links point to it. All mirrors of Dmoz. Basically where most sites starting out will be as far as number of incoming links is concerned.

Give me a minute and I'll post the updated ranking report.
If you have any ideas on what you want to test let me know.
My next test will be to add a graphic with alt text, and add meta data.

Please dont post news of this research on blogs or other forums. I want to keep control of the links so as the reults dont get skewed.

New ranking report available with changes from previous ranking report highšoghted with arrows and place change. The report is based on a search depth of 75 places.

http://www.abakus-internet-marketing.de/rreports/wmc26-05.htm

Today I'll be adding a graphic alt attribute and meta data.
Should be reindexed within a few days.

Latest ranking report..

http://www.abakus-internet-marketing.de/rreports/wmc31-05.htm

Changes (arrows up and down) reflect the changes since 26-05...
http://www.abakus-internet-marketing.de/rreports/wmc26-05.htm

No changes were made to the page so it seems to have risen on its own accord. AFAIK know, no additional links have been added, so this is all on-page. On Google.com it is already ranked 6th. With just one page and 4 links all originating from a dmoz listing. there are over 70,000 pages competing for the exact phrase "webmaster community".

There is plenty of disagreement about whether exact keywords as domain name helps or not. I think this is highlighting that it might just be an important factor and the sole reason is not just about link text in keyword domains.

From Alan Webb.

[cline]

Whether it helps or not to put your keywords in your urls is a debatable point. My take on it is that it makes site management so much easier if you do. If you're doing a logically organized, SEO-oriented site, each page is going to be associated with a main keyphrase regarding what that page is about. It sure makes sense to carry that over to the url name as it makes it much easier to mentally associate the content and the url name so you can find whatever page it is that you need to deal with. It's a whole heck of a lot better than having urls named stuff like page1.htm ... page472.htm.... Ugh.

[Terry Plank]

I don't know where the name of the file can fit into the equation.

Like others are saying, just one of the elements. I've found at Weddings By the Sea that there is a correlation to some degreee.

For example, Google search of [san francisco wedding officiant] yields #1 /links/weddingofficiantsandclergy.html and #2 /officiant.html.

Also [san francisco wedding location] yields #1 /index.html and #2 /links/sanfranciscoweddinglocation.html

This is a consistent pattern for other keywords and other search engines. At the same time, this site has been up since 1997, is a simple text based site, doesn't have "highly competitive" keywords, has somewhat decent Titles and content throughout the site, has good quality links coming in, plus other elements of a consistent strategy for search engine friendliness.

Compare this to a search at Google for [wedding favors], a significantly more competitive keyword, and it appears to be much more about the URL. Clear into result #17 before a file name /wedding_favors then not again until #32

So my experience is that it seems to be one of the factors, but certainly doesn't stand alone. Well, I don't think there is any magic bullet that, standing alone, will do the job. Like other postings above, it's an "organic" process.

[Mikkel deMib Svendsen]

I think the question of keywords in file names is one of the very good examples of how dynamic algorithms works in search engines. You can find valid examples where a keyword in the file name seems to be the dertermining factor and other examples that show it's clearly not - competitive or non-competitive. It makes analysis very confusing. Not only do we have to understand what role a parameter like this (out of hundreds) influence on rankings but also when it does so - in which searches.

In my personal experience keywords in file names are one of the less significant ranking factors. For most searches it dosen't seem to me to be a very important factor and for the few where it is I most often find it very easy to compete against - even with file names without the keyword. So, in other words, I have personally found it to mostly be a determening factor for less competitive searches.

I believe it used to matter more. Expecially in Google.

[Terry Plank]

I agree with you Mikkel, especially on "less competitive searches" comment.

Apart from ranking influence, I do like clines reference to their use for site management purposes in identifying a marketing keyword focus. Especially if there are multiple's of folks working on a site.

[dannysullivan]

Just a reminder that the Hypenated URLs (http://forums.searchenginewatch.com/forum/showthread.php?t=87) thread has a lot of good discussion along these lines. That's specifically on whether or not to use hyphens, so good this one is keeping to the issue more generally

[Carlos Chacón]

Thanks Danny, I really appreciate your comments about these lines.
Hyphens, domains names using keywords had been discussed a lot of times for the SEO purposes.
My personal opinion is that everything you can do to help the ranking of your site will be welcome. Improve, improve and improve...

[ohcho]

I recently changed all file names to include key keywords as much as possible. Though I am not sure how algorithms works, it has the weight of keywords in titles at least. What is certain is that on Google search result listing, though not in all, keywords in file names are EMPAHSIZED. That's why I know it is important. Further more, having keywords in file names makes users more likely to get interested with the pages and click! You can get more visits and more ratings from the click-through visits.

Regards.

[NFFC]

The real question is not if keywords help in file names [they do] but if they help more or less than a shorter url.

Imagine you are a search engine;

www.example.com/keyword/this-is-my-keyword-product-page-you-like-it.htm

www.example.com/237.htm

Which is likely to contain the "best" content?

[Mikkel deMib Svendsen]

> Which is likely to contain the "best" content?

This is a very interesting question. Many of you probably remember the analysis Yahoo did, I think it was last year, where they tested the quality of pages with various number of hyphens in them. As I remember they said the quality dramatically went down after two!

I am pretty sure they make similar analysis of URL-length, complexity and other such things all the time. And you are right NFFC, such testings may very well show that, generally, shorter URLs contain better content. If the analysis show that, in general, be asured it will be used. However, I have no solid proof that is in fact the case today. I do not that very long URL's hurt your but I am actually not certain about the exact cutt off.

What is certain is that on Google search result listing, though not in all, keywords in file names are EMPAHSIZED. That's why I know it is important.

You can't draw that conclusion based on what Google chose to show. What they show and highlight in a listing is not the same as how they rank pages. Ranking algo's is a million times more complex than what they show (and thats probably wise if they want to keep the majority of average users! :))

[ohcho]

>
You can't draw that conclusion based on what Google chose to show. What they show and highlight in a listing is not the same as how they rank pages. Ranking algo's is a million times more complex than what they show (and thats probably wise if they want to keep the majority of average users! :))

It seems that Google does use them! One of my web-page was ranked at about 170s on a keyword set search. After I change the page file name with the keywords and redirecting the old name to new, it went up to about 90s immediately!

After all, even if your pages appear in front, unless users pick your pages, your page ratings will quickly slip away! Making users interested on your listed pages will be also equally important. Having file names suggesting contents is a good tactic to make them interested.

Regards.

[Mikkel deMib Svendsen]

All I can say is that you draw way to broad and general conclusions based on very limited data :)

[glengara]

FWIW, I've an old site designed on the themed pyramid which uses KWs in file names.
On-page they targeted "SEO widget", but the file names used "search engine optimization widget".

Anyway, about 6 months ago these pages also started ranking well both in the normal SERPs and in Alllinanchor for "search engine optimization widget" terms, but since the update they've disappeared from both, while remaining steady for "SEO widget" terms.

My conclusion was G had been using file names in their allinanchor calculations, but have recently stopped doing so.

http://forums.searchenginewatch.com/showthread.php?t=6087

[bdw]

As someone mentioned earlier file names are a highly logical way for the search engines to determine what a page is about but I have been wondering if Google is using an over optimisation filter that could be tripped by too high an incidence of the same phrase or word in all the strategic places.

For example page name = fuzzy_blue_widgets.htm, page title = Fuzzy blue widgets and H1 = Fuzzy blue widgets. While these elements could be selected "naturally" it is also likely that if they are all exactly the same then this has been done for SEO purposes and may result in a loss of ranking points.

I launched a hobby site about a famous person in November last year. It had no commercial content or purpose but it was firmly sandboxed. A month or two ago I started varying the three elements I mentioned above. Let's say that I had a page on my site called john_doe_bio.htm with page title and H1 the same. I tried leaving the page name as was and varying the page name/header combo to something like doe-bio and johns_bio.

The effect has been that the site jumped out of the sandbox last month and it now gets found in position 27 in Google for the person's name. This is not spectacular but there are over 7M results for this search and previously it was not in the top 1000. I also think that a further increase is likely over the next couple of updates. This is slightly complicated by the fact that these changes coincided with the Bourbon update but my (non-scientifically determined) conclusion is that page name is a reasonably important factor that cannot be ignored but don't overdo it elsewhere. ;)

[ohcho]

The most important factor in Google SE rating is determined by SE users. Not by Google nor by SEOs. If you have a web-site containing pages that are popular to SE users, your web-site and pages will get highly ranked in time. To be successful, you have to consider full cycle as follows;

search -> click -> bookmark

If you are tinkering with web-pagse that nobody wants to read and bookmark, SE users will know from the output listing immediately and will ignore your pages. You may be able to bring your pages ahead doing SEO. But it will be temporary. SE users will skip over your pages and as a consequence, your pages will slip away very quickly!

It may be argued about SE use of file names. But at least SE users use file names filtering bogus pages from search result listings. It is very important to have file names that carry information!

[Mikkel deMib Svendsen]

The most important factor in Google SE rating is determined by SE users.

Thats certainly not my experience. There are many factors but what users clicks on seems to be one of the least important. Sure, Engines tracks it but it is not what makes webpages rank well and stay there by the end of the day.

[ohcho]

Last September, we had less than 50 SE visitors, mostly by product name searches. Now we have a few thousands a month. This number is keep growing thanks to popularity and *high click-through rates*. They are almost from Google. Now most key keywords are within top 10 ranked in Google.

If your rating is too low, then you can tinker with other factors. Once you get some ratings with inbound references from a few high rated sites, the rest is depend on what and how you write, eventually voted by SE visitors. If your keywords are heavily congested, then it will take effort to improve ratings.

Note that Google prospers by bring web-pages what SE users seek after! That is, it brings most relevent information to seekers. If you understand this and develop web-contents accordingly, Google SE will reward your effort. I enjoy that reward now! But I don't know how I can repeat the same with Yahoo and MSN.

Regards.

[bdw]

You may be able to bring your pages ahead doing SEO. But it will be temporary.

Really? I think not.

[pleeker]

ohcho, what Mikkel said several posts ago is 100% accurate. You're drawing too many broad conclusions from a very limited data set.

Yes, it's quite likely that when click-thru is being measured Google keeps that data and does something with it. There's an ongoing project on the SEM2 group where people have been asked to click on a certain page in the SERPs to see if they could move the page up the rankings. Last I checked the thread, that page had seen little or no benefit from these clicks -- there aren't enough people clicking, in all likelihood. It would take a staggering amount of clicks for that one piece of the algo to outweigh all other factors that determine ranking.

Using click-thrus as a major ranking factor has been tried in the past (see Direct Hit) and deemed a failure because it's too easy to be manipulated. As SEs move toward more personalization, however, it seems likely that your personal clicks will help determine your personal SERPs. But it's unlikely that click-thru will ever be a large factor in determining general ranking.

[JEC]

You may be able to bring your pages ahead doing SEO. But it will be temporary. SE users will skip over your pages and as a consequence, your pages will slip away very quickly!
I haven't been in the SEO arena for too long so my comments may be off. However, I can tell you that I started doing the SEO thing for our company about 5 months ago and one of the very things I did was to incorporate our keywords into links, and into some filenames where appropriate. We were up in the 100's before, but after a while we came down and we now float around page 2 or 3 on the top SE's for those keywords.

I think the whole idea of SEO is to use practical techniques that get noticed by the SE's and add relavancy to your site for the long haul. You may see your pages slip here and there, but it shouldn't be a large slip unless you are practicing some voodoo the SE's really don't care for. After all, if your competitors are also doing SEO, wouldn't their sites move up as well? We can't all be at the top. I also don't think that all slippage is due to poor SEO techniques as much as it is the SE's changing the way they look at your pages. IMHO, incorporating the keywords into filenames helped us move up. It wasn't the only thing that worked for us, but I think it certainly added to the equation.

JEC