[thelist] More on Search Engines

[James]

Russ wrote:
>> Whatever an all-too-clever "SEO" can implement, sooner or
>> later, a spyder can be written to detect.  If a browser can
>> interpret it, a spyder can be written to do likewise.
>> Google's bread and butter depends upon their being at least
>> as clever as the SEOs.  The SEOs that I have faith in are the
>> ones that work within, and endorse, Google's guidelines.
>
> It doesn't truly appear that anyone fully is aware of Google's
> guidelines; there are questions popping up all over the place
> regarding this.  There are a lot of forums on SEO that have many, many
> professionals very confused and concerned with how to deliver results
> to their clients, so it's difficult to believe that these guidelines
> are entirely commonplace and/or that they're endorsed.

There are, of course, SEOs that work "within the law" and those that work
"outside the law" - just as there are 'get rich quick' schemes that are
legal and those that are not so legal.  IMHO Ed's comment that "whatever an
SEO can think of, a bot can be written to detect" is spot on - it's just the
case that most SE don't have the resources to chase the 1% (?) of webmasters
who cheat.

As soon as tricks become mainstream (like the old <meta keyword> stuffing
trick), the search engines are forced to act (which is why very few SEs
actually take notice of the <meta keyword> tag these days). So, why not pick
the SEO who understands and works within the SE's guidelines?  It certainly
won't get your site to #1 quite as fast, or quite as often, as the
unscrupulous SEOs, but it's *far* less likely to get your domain banned
completely for sharp practice.

> Further, following one search engine would seem about as crazy
> as following one browser version.
> A spider might be able to written to go through and read CSS and see
> how it's applied, but how is that going to affect the speed of a
> spider and the lag in which it takes to spider sites?  I don't know
> the answer to this, but anything that can be written to do more
> typically results in it taking more time to complete. [snip[

AFAIK no mainstream spiders request and/or parse CSS.  For a start, how
would search engines deal with sites with robots.txt saying

User-agent: *
Disallow: /css

or sites which rewrite requests for CSS files for Googlebot using
mod_rewrite?  Also, there are so many glitches/quirks in client's rendering
of CSS that a bot could never be 100% sure what effect a particular html+CSS
combination would have, and therefore whether it was 'bad' or not.

Of course, I could be wrong...  many of the sites I compete with use hidden
divs stuffed with keywords and links to increase their rankings.  <sigh> If
only I could report them all to Google in batch mode :-(


James