"Dynamic" Pages

I see various, sometimes contradictory statements about whether and how Google (and the other engines too) index "dynamic" pages. I am not even sure exactly what a "dynamic" page is. I do know that Google (and I am pretty sure Yahoo) handles shtml satisfactorily (I have scarcely a page that isn't shtml).

What I am unsure about is pages created on the fly by php. The Backlink Tracker reports 1850 pages in my URL. My own count, which is approximate but close, shows about 1520 page files. But if links to php-created pages were included, there would be literally thousands more. (I have 1242 pages which are each for a given book title; each such page has links to several--usually quite a few, sometimes many, many--php URLs that are individual-edition pages created on the fly by php using Amazon's AWS.)

My guess is that Google is not looking at those php-created pages, but can anyone tell me for sure how Google treats such things? (If it matters, the calls send only 2 parameters.)

Side questions: do either <form> posts links or <img src=> links "count" with Google as outlinks?

 

Does the site page count returned by the Tracker reflect Google's understanding of the size of the site, or is it calculated independently?

I ask because the numbers perplex me. I have about 1520 static pages (well, shtml, but static in this context). Of those, 832 are pages with an absolute minimum of one unique dynamic link, and most have at least several such (one alone has hundreds). The Tracker returns me a count of 1850 pages: that's well over the static total, but vastly under the dynamic total. (OK, I did a run to check unique dynamic URLs: I got 8599, but in fact some are not unique, since a few books--including the most-linked one I list--have alternate titles; even so, though, there are clearly many thousands of unique dynamic URLs there.)

Just for fun, I ran one of the dynamic URLs as a Google search, and got back nada.

Does it matter that the pages containing these dynamic URLs are themselves built anew once a day? (That is, does Google not have time to run the dynamic URLs before, from their point of view, the page has changed--even though the actual change is usually little or none.)

I suspect that effectively multiplying the page count of my site by a factor of 5 would not be a bad thing for PR and ser. I reckon I need to sort out mod_rewrite.

 

I dont quite understand this - what do you mean they are 'built anew once a day'? And what does 'run the dynamic URLs' mean?

All the articles on both my sites below are 'dynamic'.

As an example, the mortgages page on my House Buying site doesnt actually exist. The URL is '/buying/mortgages.html', but 'buying' is not a folder, it is a PHP script which extracts 'mortgages' from the URL, queries the database for articles with that title, then displays the appropriate article.

 

My site is about books. I recommend, at present, 1242 titles (a few of which are duplicates, books published under multiple titles) of a particular sort. Every night, automated software uses Amazon's AWS to obtain data about all editions of a given title and rebuilds a static page for that particular title listing all available editions, from all six Amazon national divisions, with their prices and availabilities; even if those data don't change daily, the various currency equivalents for the prices (which I also give) will almost certainly change daily, owing to daily fluctuations in the exchange rates, so most or all of those 1242 static pages change daily.

Each of those 1242 pages has in turn links to a php page that will return the instantaneous price and availability being reported by this or that Amazon division for that particular edition. I checked more closely last night: I have 6812 unique calls to that php page; if those were each treated as a page of my site, my page count would go from 1520 to 8332.

I am told that Google will slurp up such pages, subject to a couple of caveats. I wonder if I am caught in one of those caveats: one of the two parms passed is "myid=sometext", and I wonder if Google might be seeing that as an "id=" parm, and thus ignoring the URL. I am going to try changing the parm's name in tonight's run, and we'll see. But, unless that produces results, I reckon I'm going to have to figure out mod_rewrite.

 

thank you for this tidbit, ive been wondering why yahoo + google have been having trouble with spidering my deeper links, i had the id variable

-dan

 

Sometimes you can get away with 'ID' in the url, so long as it has only a couple of numbersafter it. I have a directory and the ID numbers are not too bad, (still not happy with them though, should re-write) Also the number of query strings will delay/prevent spidering, 2 is ok 3 is pushing it, 4 is right out (as the famous holy handgrenade of Antioch sketch goes)

 

I'm happy with dynamic pages . We are running a travel related site which consits to 50% of dynamic pages with two vars (airport & service). These pages do not exist but all 200 are indexed and reasonable positioned.
The "trick" maybe that each "page" is at least twice referenced from "real" pages and from a fairly massive site map.
They are also in Y and I.

I have thought about rewrite, but G's crawls being a bit more forgiving when hitting dynamic page is actually quite OK as we sometimes have up to 14 crawlers on the same site all hitting dynamic pages.

I also (sometimes) think they watch each other as Y and I are usually just finishing their crawl and G takes over or they come in at the tail end of G's crawl.
M

 

Thanks for that tip, google has now indexed a bunch of my pages!

But will google pass on PR ok on these dynamic pages?

 

Yep, no worries. PR is passed whatever the receiving page is (even if there is nothing there!).

 

I write most of my sites in php. I found out that allthough PR for my dynamic pages is listed as zero they seem to have a real higher page rank then zero.

I experimented with a site Bali Hotels to make some dynamic urls appear as static ones. Now the dynamic pages show pagerank. Together with some other structuring I did, it al seems more effective to have static (appearing) pages. Allthough the system applied is not totally listed in google yet, it doesn't yet seems to differ in total ranking of the site for general keyword ranking, it seems to matter for the ranking of the individual items on the site.

Please note that I don't have mod_rewrite on my server, but the system applied to the above site works as well.

 

I've built a tool which spiders an entire site and creates what I call a "Static Shell" - basically a static html version of your website. Just put a link to this shell and voila, search engines will now index your content. The benefit of this method is that you dont have to install anything on the server, or implement crafty programming. Also, you are guaranteed of getting indexed by all major SE's.

The tool works great on dynamic sites that aren't getting completely indexed. I've used the tool for some clients with large e-commerce shops and it has worked wonders for them.

Contact me at farum21@hotmail.com if your interested in using this tool.

 

Since we’re on the topic of dynamically loaded web pages, I would like to ask a SEO question. I’ve seen all these recommendations regarding to limit 1 to 3 keywords per page when optimizing a website. How do you optimize the metatags, keywords and title of a dynamic website for various topics?

Most, and I admit its not many, dynamically loaded websites that I have reviewed, have one page and based on how a user navigates through a table of contents or menu, the main body of the page is filled with data from a database.

I’ve seen header tags <H1>, <H2>, ….<H6> used effectively but I have never seen the metatags, title, keywords change or updated as the page is loaded with information.

Is it possible to optimize the header area of a dynamically loaded or database driven website like you can a static website?

 

My site is dynamically generated and I've written the code such that each page has it's own title,keywords and description tags. There is a default in case I don't enter one in the admin utility when I'm adding a "page". This has proven very effective in gaining very high serps for a very wide range of key words. I've since re-written the application and made it into something that I am close to releasing commercially. It's called CMS Writer and if your interested, drop me a line. I"m going to be looking for some beta testers to hammer on it and make sure it's ready for prime time.

 

I load metas into the database driving sites thus I use a db call per pageand something like
// see if we have decent titles etc
if (empty($mtitle)) {$mtitle = "$site_name";}
if (empty($mdesc)) {$mdesc = "$site_desc";}
if (empty($mkey)) {$mkey = "$site_key";}
?>
<head>
<title><?echo "$mtitle";?></title>
<meta name="description" content="<?php echo $mdesc ?>"> <meta name="keywords" content="<?php echo $mkey ?>">

Thus I can alter / build them without touching the pages.
M

PS On high traffic sites I tend to use mod_rewrite and caching....