characters, only about how the engines parse them.
The sections to be looked at are:
1) the 'depth': how many apparent sub-dirs down from the root (this is part
of the 'long URL' definition)
2) the 'name': actual name of the resource requested, if any (such as in
your example)
3) the query string: length of query string and how any vars are to be
interpreted (this is the other part of the 'long URL' definition)
In your example you had a long file name in the URL but that is it. You
weren't deep, you didn't have a query string, and you report the name of the
resource as corresponding in context to the content of the resource once
spidered. Congrats! You are doing exactly what the search engine spiders
like to get high positioning. (p.s. it would also help if you repeat that
info in the <title> tag as well as in a <h1> near the top of the page.
the 'depth' and query string are more problematic. It's been clearly stated
by search engine companies that they do look at the depth of an URL, and in
fact most won't bother giving any additional weight to URLs with depths
greater than 2 or 3, simply because of the 'techie' solution a few years
ago of addressing the query string problems by creating what they called
"search engine safe" by substituting slashes for the ampersands and equal
signs on the query string. That is, since a "deep" URL often represents what
is just a regular URL with a long query string, they just ignore the effect
altogether. You still get *in* the engine when spidered, but you dont rank
nearly as high as if you had a 'shallow' URL
query string: This has been the most uneven set of solutions addressed by
the search engines. Some of them parse it completely nowadays. Most of the
older ones don't do anything with it. The smarter search engines, like
Google, actually have some sense of certain words to look for that indicate
a 'dynamic' site (to most CF'ers, a good example would be Fusebox apps --
Google knows that "fuseaction" is a common way of distinguishing what could
otherwise be a resource name) But this doesn't always help because
sometimes it is a combination of query string params that actually
distinguish one resource from another:
method=getProductDetails&ProductID=123
is different than
method=getProductDetails&ProductID=456
whereas
method=HomePage&DayOfYear=2
is likely not very differnt from
method=HomePage&DayOfYear=3
versus
method=AboutUs&CFID=123456789&CFTOKEN=987654321
where the CFID/CFTOKEN represent nothing at all about the site since they
are only a stateholder
And anyway the spiders have to be programmed to care (or not) about such
differences (such as in the case of the 'fuseaction', etc).
All in all, dynamic URLS should be mapped to static ones, and this mapping
should be be done by your marketing department
One last comment, since this always seems to get lost in a techie discussion
about search engines. Placement in search engines is about marketing, not
about technology. And because it's about marketing, it will often look to
techies like "oh that's just a trick to get high placement" (the implication
being that if they learn all the tricks then they will be search engine
marketers too).
I know we all agree that the bar for calling oneself a marketer is
extraordinarily low; nevertheless, we should keep in mind that just as we
would laugh at marketers who presumed to explain proper use of CFCs, so too
do marketers laugh at us when we attempt to perform their job. It's good for
us to be aware that when we're tweaking our sites for search engine
placement that we are wearing a different hat than we normally do and that
we should be aware that we are likely out of our realm of competent
experience -- and to therefore tread lightly.
----- Original Message -----
From: "Matt Robertson" <[EMAIL PROTECTED]>
To: "CF-Talk" <[EMAIL PROTECTED]>
Sent: Friday, January 02, 2004 12:32 PM
Subject: RE: CMS Solutions (Friendly URL's)
> Joseph Flanigan wrote:
> >Your statement is interesting. Do you have some references that support
>
> >your statement?
>
> >>One needs short relevant file
> >>names in a short shallow URL corresponding to actual content in the
> >>document to make the biggest impact in the search engine rankings.
>
> We could analyze this forever, but lets ask whats a long url? The part
> about the relevant keywords is critical of course; that's common SEO
> advice. Take this url for example:
>
> http://tamsenmunger.com/Nicolas_Trudgian_The_Checkertail_Clan.cfm
>
> Google 'Checkertail' only. #1 there and at Yahoo (as it also is for
> other relevant keyword combos). The page (and the site) is about 5
> weeks old. I'd call this a long url, but we wanted both author and
> product name in there.
>
> I've got similar results on 1, 2 and 3-word combos where all seem to
> handle the 'long' url's just fine on this engine.
>
> Repeat: on this engine (Google/Froogle/Yahoo). Go to MSN and, at
> present at least, these pages aren't ranked at all. Their urls for this
> site are outdated by months so maybe this will change. I can live with
> free #1's on just these engines, though.
>
> --------------------------------------------
> Matt Robertson [EMAIL PROTECTED]
> MSB Designs, Inc. http://mysecretbase.com
>
[Todays Threads] [This Message] [Subscription] [Fast Unsubscribe] [User Settings]
