View Full Version : Is "Dynamic Site Caching" ok?
Jeff Martin
08-18-2004, 06:14 PM
A search firm that says it specializes in the search engine visibility of dynamc sites says they use "dynamic site caching". I have visited a site that uses this firms technology and I wanted to get some feedback because a client of mine is considering this.
When a "site:" search is done, I would say about 90% of the visbile results have static URLs. I only see a few dynamics. When I visit a static page from Gs (site: search) SERPS, the page I get has a <noscript> with links in it right after the <body> and the links on the page all have static destinations. The page visually looks normal. As far as I can tell the content of the static pages is the same as the dynamic, EXCEPT when I visit the homepage Im sent to a page that is slitley different than the home page a user sees when just typing in the domain name.
If I visit the site by just typing in the domain name I see a dynamic site with all destination links being dynamic and the pages do not have <noscript> links in them. One exception to that is the homepage. As I said the homepage I visted when clicking an internal link from a static page is slightly different than the one I see when I type the domain name in and am looking at all dynamic URLs.
Im not all that familair with "dynamic site caching" can someone offer some insite? Is this an automated URL rewrite or a type of cloak with an index of duplicated (with enhancements) static pages? Possible repercussions from the SEs? It looks like these static pages are fed to the SEs only.
Gracias!
rustybrick
08-18-2004, 06:29 PM
I would say that one of my firms specialties are search engine visibility of dynamc sites. I see no real reason to deploy "dynamic site caching" to increase visibility. What we do is a form of URL rewriting (i.e. mod_rewrite in conjunction with PHP and MySQL). Again, I see no reason for deploying "dynamic site caching" unless your looking to save bandwidth and database calls.
We cache a lot of pages in order not to speed up the site and reduce server load. All it does is makes a static version of a dynamic page for a specific time period. When it comes to e-commerce sites, with many goods, a homepage is likely to be cached. This way the images and products on the homepage are fast to load (they do not have to request data from a database to generate the page).
Hope this clears things up. If anyone has more to add, I would be very interesting in hearing.
Jeff Martin
08-18-2004, 06:49 PM
So is this an IP based solution that redirects spiders to static pages (or users to dynamic ones)? How are these pages cached and fed to the SEs?
Im a little concerned about the home pages being slightly off and the <no script> tags with the links in them.
rustybrick
08-18-2004, 06:57 PM
To tell you the truth, I am not sure what they mean by "dynamic site caching" as it relates to search engine visibility. It can really mean a dozen things.
All I have done is re-write the URLs. I'll give you a real life example.
A recent site we put up is purely dynamic, the company name is Smart Tuxedos (http://www.smarttuxedo.com/) (give them some anchor text for allowing me to use them as an example). If you click on the "dynamic" links (Tuxedos, Dinner Jackets, Shirts, Vests, etc.) or any of the products links, you will see a URL that looks like http://www.smarttuxedo.com/Tuxedos-1/. We feed a number in the URL (1 in the case for Tuxedo) as well as the keywords associated with that page (not that I feel keywords in the URL help). The id is what we use to tell the code what data to bring back from the database (explaining it in very simple terms).
If I had 100,000 visitors to this page in 30 minutes, then I would consider "caching" this page. What that means is I create a static version of the page. This cached page does not require me to call the database for the necessary data (i.e. product name, price, description, etc.), it stored a local static (cached) copy of this page. That makes it quicker to download and saves server resources.
Now to use caching alone to improve search engine visibility? I don't see why it would be beneficial. Maybe they are creating search engine only pages will a caching technique. If so, then call it what you will - it sounds a bit border line to me.
So there is no IP based redirects going on with Smart Tuxedos, the users surf the site just like a search engine would.
Jeff Martin
08-18-2004, 07:10 PM
The difference in the URLs is this:
The tuxedo site uses a URL rewrite. A visitor and SE see the same thing.
In my example the site is dynamic to every user who visits www.sitename.com. All the links are dynamic.
BUT, when you look at the pages in the SERPS they are all static. When you visit the pages from the SERPs all of the links on all the pages are static. This doesnt seem like a rewrite to me but a second set of pages (which contain some SEO underwriting).
So we have two sets of pages. One for the SEs and one for users, however the cached static pages tailored for the SEs (no copy changes, only HTML changes) are visually indentical to the dynamic ones.
I agree, it sounds borderline. Anyone elese want to chime in?
rustybrick
08-18-2004, 07:25 PM
Jeff, I would not sign up for that then. It can be done in a fashion that you have one page for the SE and the same page for the user.
Jeff Martin
08-18-2004, 07:30 PM
It can be done in a fashion that you have one page for the SE and the same page for the user.
Sure with a mod rewrite. For windows IIS boxes its a tad more difficult than Apache. My client is nervous about having a rewrite run in the process of IIS and not knowing how to use regular expressions. They want the ability to be independant and not pose any risks to there environment. While I know URL rewrites are safe and the proper way to go, the client just doesnt feel the same way.
Mikkel deMib Svendsen
08-18-2004, 07:51 PM
This is what is usually refered to as "content replication". In my book this is one out of a number of available solutions to get a dynamic website indexed. But, one of the last options I would normally recomend. Fixing the site, or applying some kind of URL-rewrite is most often better solutions but if for some reasons those and other better solutions may not work for you then site replication is definately a valid option.
However, with site replication you have one large problem tp deal with: Inbound links. If the URL's users see is not the same the engines index you will not get any link value to the right pages
rustybrick
08-18-2004, 07:52 PM
Understood. I feel less comfortable building one page for an spider and the duplicate page for a user. That is me.
Maybe there are some good Windows people out there?
Mikkel deMib Svendsen
08-18-2004, 08:01 PM
I don 't think it's the best option either but my job is usually to find solutions - and often within very narrow limits. Therefore I find it usefull to have as many different options available even though some are much better than others. Site replication is still better than having no indexing at all - if that's the alternative.
Jeff Martin
08-18-2004, 08:03 PM
However, with site replication you have one large problem tp deal with: Inbound links. If the URL's users see is not the same the engines index you will not get any link value to the right pages
Excellent point Mikkel. Sure enough, a quick look shows the static pages have a PR0.