Ted On Flash: Flex Directory, SEO, and Flex

Flex Directory, SEO, and Flex

DIGG IT! 5 Comments Published Tuesday, August 07, 2007 at 6:02 AM .

I did some modifications to the Flex Directory for SEO optimization based on Josh Tynjala's article. One of the issues of search engines indexing XML documents is the only schema they understand is html. This legacy is a blessing and a curse for structured data. HTML can be XML compliant but the reverse is not always true.

The Flex Directory is a "Model First" approach to RIA. The model (raw data) arrives before the presentation layer or "view" of the application. What is interesting is that spiders are looking for the model not the view of an application. Spiders do not care what an application looks like, they are interested in the data on the page, not the surrounding markup. Right now the directory model just contains information about the company listings but over time I will be moving all data in the Flex Directory over to the XML format including the strings within the application. To expose this data to SEO, it makes sense to support the html schema in parts of the XML document allowing spiders to process the data into the search engine listings without issue.

Here are some changes I made to the Flex Directory for SEO:

1. Added sitemap.xml file
2. Added robots.txt file
3. Changed root element from "directory" to "html"
4. Added header element and inner title element
5. Added meta "description" and "keyword" elements

I didn't include a body (yet) as I want to see how this document is processed in the search results over time. I will be tweaking things ongoing to try to get the number 1 spot on the search engines for the Flex Directory. What will be interesting is to see if the spiders ignore the xml-stylesheet and process the document as html. If that is the case, then the results page on both Yahoo and Google should change.

Special thanks to Josh Tynjala for his article on SEO and RIA. The article is a great read and he dissected the Flex Directory in the article. I took some ideas from his post and this influenced the changes to the directory as noted above.

Thanks Josh, great post!

Regards,

Ted :)

5 Responses to “Flex Directory, SEO, and Flex”

# Josh Tynjala August 7, 2007 9:45 AM

Thanks for three links back, Ted. The only thing missing is a giant image of some sort. ;)
# robdixon August 8, 2007 12:16 PM

Hi Ted,

That's a brilliant idea to use XSL to load the Flex app for the underlying XML content.

I know you're waiting to see how the search engines pick up your recent changes before tweaking things more. I think the following changes could have a big impact going forward:

* Adding an introductory paragraph in the XML/HTML that describes the application. You might not even have to display this paragraph in the app itself, though it could be useful for new users if you did. Adding more text improves your "text to code ratio" as well, which is a Google factor.

* Using a more descriptive file name for the HTML page, like
flex-consulting-directory.html, should have a big impact.

* Search engines like pages that have hyperlinks in and out. The XMl has plenty of attributes like url="www.domain.com". I wonder if those are counted as hyperlinks? Or do you need to have it formatted with the http:// in front? Or with an actual anchor tag? On the other hand there's some speculation that "link farm" pages are downgraded by some search engines, which is another reason to have more text on the page.

Anyway, great work.
# Jim Phelan August 8, 2007 12:33 PM

Hey Ted,

Not really related to SEO, but is there any way to edit a listing? We (Stream57) are a premium solution partner now, and it might be useful for others as well.
# Scott Lepich August 10, 2007 10:05 AM

I'm not a huge fan of xml any longer, but just for the sake of discussion: why would you not just use xhtml as your schema? It would seem to be the best intersection of html and xml. It works for non-flex sites out there so I'm sure it would be adequate to describe your data set. Why not use a microformat rather than all of these xml attributes?

<company name="Outsmart" contact="mitch.olson@getoutsmart.com" url="www.getoutsmart.com" partner="1" country="NZ" location="Auckland NZ" developers="12"/>

Just thoughts.
# Eric January 11, 2008 1:51 PM

@Scott - If prefer to use HTML, space150's "Faust" technique might be a good solution. Instead of grabbing XML from the page, pull XHTML and Microformats. This is basically "Progressive Enhancement", where the Flash "augments" your HTML experience. It's the same data model (HTML) either way.

http://blog.space150.com/2007/1/11/faust-flash-augmenting-standards

Stuff I am building:

Ted Patrick - Events & Community @ Adobe Systems

Note: This is the personal blog of Ted Patrick. The opinions and statements voiced here are my own.

Flex Directory, SEO, and Flex

5 Responses to “Flex Directory, SEO, and Flex”

Where to find me:

Latest

Lists

Links

Jobs

Archives