>increasingly annoyed with SEO specialists who make it all
>out to be this mysterious black box.

Much of SEO is actually very straightforward, once you've expended the effort to understand what the search engines want. The value of a good SEO lies in the epxerience to understand, quickly, what needs to be done, and what the best methods of doing it would be. That sort of experience is quite "black box" as it relies heavily on human judgement and long experience

>> >Things that seem help search engines index dynamic content

1) Fewest number of parameter / value pairs (example.com?parameter=value&parameter2=value2) possible. If some sort of URL rewriting process is possible, this can be reduced to zero, which is ideal.
2) Link popularity / PR score. Google has specialised robots tasked with the spidering of dynamic sites. These resources are relatively scarce, so priority in frequency and depth of crawl are given to sites that are more popular (get them PR8 links!)
3) Freshness of content. Sites the refresh (or at least seem to refresh) their content more frequently are spidered more frequently. This applies to dynamic sites too
4) Clear navigation. Some dynamic sites can create infinite loops, trapping a spider in an unbreakable series of links. If these are detected, the site will not be spidered, and will therefore drop out of the index

Dynamic sites CAN get indexed, and rank well. Microsoft.com have just under 600k pages in Gogole at the moment, and the majority are dynamic pages. You just have to know what the search engines want, and how to present it to them - back to good old human judgement, here....

Thanks - this makes sense. Have search engines been public about this or is it what SEO experts have figured out with experience?

Lawrence

On 13:08:51 28 April 2004 TallTroll wrote:
>On 11:57:48 28 April 2004 Loz wrote:
>>I wonder if search engines are able to distinguish
>between
>>accessible and non-accessible code. Not necessarily to
>>penalize sites that are not, but to have a filter in
>place
>>so that users can search only sites that are
>accessible.
>
>Search engines don't distinguish as such, but many of
>their ranking criteria are related to accessibility
>issues. All other things being equal, a more accessible
>page will always outrank a less accessible one, as it will
>score more highly on several minor parts of the serach
>angines algorithms.
>
>For instance, good descriptive alt and title attributes
>help users using screen readers understand you site - and
>help you rank by providing extra places for you to place
>keywords. Sites that validate also get a small, but
>significant boost in ranking. Sites that make logical use
>of of the structural elemts of HTML, and use CSS to
>control layout and presentation (instaed of dodgy
><table> based layouts) will also score on relevancy,
>and page size, and text-to-code ratio.
>
>Do you see where I'm going with this?

There are also ways to 'hide' dynamic URLs with static looking ones. An example would be this page itself (I think).

Lawrence

On 12:14:10 28 April 2004 Obi Felten wrote:
>Very happy to see a pragmatic approach, I am getting
>increasingly annoyed with SEO specialists who make it all
>out to be this mysterious black box.
>
>One thing not addressed here yet is dynamic product
>content & links: On large retail websites the majority
>of content is product content, which is dynamically served
>by their commerce engine. Category and product links are
>generated based on what is in the product database, and
>the links don't contain the category name but stuff like
>catID=20 to keep the URL length down. There are typically
>few pages of flat html content, e.g. product guides (which
>do help their Google rating apparently, even if no
>customer ever looks at them).
>
>Things that seem help search engines index dynamic content
>(I'm no techy, so this is perhaps put too crudely, hope I
>don't get anything wrong!):
>1) Page title: Add the category/product name into the page
>title for category/product pages
>2) Unique URLS: Don't put the identifying category ID into
>a query string at the end of the URL after.html or .jsp
>where the crawler ignores it. www.example.com/category.htm-
>l?catID=20 looks the same to the crawler as
>www.example.com/category.html?catID=45, so the search
>engine will only index a single category page on the
>entire site - which may be your most important category,
>or a tiny unimportant one.
>For example Amazon's URLs are all unique.
>3) Cache the whole page as HTML, fixing the links to
>categories and products rather than generating them
>dynamically whenever the page loads: This means you can't
>personalise the content, but I would argue being found by
>search engines is more important, and it will also speed
>up the page load time. I think Comet do this, which may be
>one of the reasons why they come up very high on Google
>for generic terms such as 'washing machine'.
>
>Am I getting this right? Any others? Would be interested
>to hear.
>
>Obi

I think it's a good idea nevertheless as URLs become easier to 'understand' and remember. Which is the idea behind a service such as http://tinyurl.com/.

Lawrence

On 14:36:34 28 April 2004 Ashley wrote:
>[NB The whole forum / SEO / mod rewrite / dynamic URL
>discussion has been had in depth before - read the thread
>starting at http://www.e-consultancy.com/forum/100305-
>forums-and-search-engines.html ]
>
>I wouldn't call it 'hiding' but, yes, we do use a mod
>rewrite to "flatten out" our query string URLs
>to become 'static' pages. When we initially did this about
>6 months ago or so it did make a big difference to our
>traffic as more pages crawled. However, since then the
>spiders have got clevere and (I understand) can crawl
>query strings with up to 2 or 3 parameter values in them.
>
>So you don't *have* to do it and I'm not even convinced it
>makes much difference to have the search term in the URL
>itself but, hey, it hasn't done any harm either.

Rant on Alex, I say.

Here's a thing. I often click on Google's Adwords - mainly the lower ranking ones - and usually when the natural results look too polluted even to bother with. This is because I tend to find that if a company is prepared to pay to advertise, they are also sensible enough to give me some decent content too - in other words, they have invested in the site.

Would be interested to know anyone else's views?

John Duffy