Subscribe to RSS
get email updates
polyGeek.com
polyGeek.com polyGeek.com

Taking Flex SEO for a test drive

July 1st, 2008 . by polygeek

Unless you’ve been hiding under a MSN search box you’ve heard the news that Adobe has worked with Google and Yahoo to make SWF files truly indexable by search engine spiders.

I’ve put together a quick Flex application this morning to take this new feature for a test drive. It’s just a TabNavigator displaying a few tabs of content about Greek gods. My goal isn’t to become the number one result for zeus. I just want to see the content get indexed and see what the search results look like.

a few greek gods

If you would like to help out with this little experiment then try linking to the application so that it might get crawled sooner. Wouldn’t it be cool if this little experiment ended up being on Google’s first page for “greek gods zeus”? I’ll do a site-search every few days to see when it gets indexed and report back in another post when it does.

I pulled the content off of the Wikipedia.org site and dropped it into a database here at polyGeek.com. Then I used AMFPHP to grab the data and create as many tabs in the TabNavigator as there are results. And I added toolTips to the images if you mouse over them.

The nice thing about the TabNavigator is that the BrowserManager is automatically engaged. And I wrote 3 lines of code to enable the deepLinking in the app so you can land directly on a specific tab depending on the URL. And as always, mi coda, su coda.

In the code I’ve added the SQL, images, and my GoRemote.mxml to help you recreate this if you want. You’ll just need to get AMFPHP going yourself and add my AMFdata.php file to the ../services/amfphp/ folder. I might cover all the more in depth in a later post, someday.

Addendum: I just read the blog post at Google and it mentions that

Googlebot does not execute some types of JavaScript. So if your web page loads a Flash file via JavaScript, Google may not be aware of that Flash file…

I’m using the default Javascript/HTML code that FlexBuilder creates. So I’ll see if it gets indexed. And if you’re wondering why I use the default FB code it’s because I’m too lazy to take the additional step of using SWFobject, most of the time. :-)

Post to Twitter Post to Delicious Post to Facebook Post to Reddit Post to StumbleUpon


similar posts

18 Responses to “Taking Flex SEO for a test drive”


comment number 1 by: brett

You might want to think about using SWFAddress so you can get nice deep link URLs, instead of the normal nasty Flex links.


[...] aquilo que o programador pode fazer para obter o melhor partido destas novas alterações, as experiências da comunidade podem ser uma boa ajuda, no entanto as antigas técnicas continuam válidas (deeplinking, sitemaps, [...]

comment number 3 by: Tim

I was considering a test like this. I was going to use Google’s sitemap feature to promote the indexing of the test SWF though. If you haven’t used it before check it out here:
http://www.google.com/webmasters/

There’s an XML format you can follow which specifies which links Google should index. This might also let you know when Google has indexed it since you can login to your account and check the status.

comment number 4 by: Savvas Malamas

“In Greek and Roman mythology, Apollo is one of the most important and many-sided of the Olympian deities.”

“On the 2007-*’s desktop, Apollo is the most beautiful and exciting technology and as a Greek I am very proud that it had this name even for the alpha phase”. Thank you Macromedia, thank you Adobe!

comment number 5 by: polygeek

@Brett, I haven’t used SWFAddress. I’ll check into it.

@Tim, I’ve heard of Google’s sitemap. I was even using a Wordpress plugin that created one for this site but it conflicted with one of the Wordpress updates. I should re-enable it.

@ Savvas, you’re such a geek, I mean Greek. :-)

comment number 6 by: Brian

Isn’t it a problem that all the gods will be indexed under the same URL? It seems to me that the new SEO feature really makes deep-linking impossible for SEO, because each deep link into the app will be indexed using all the reachable text (i.e. everything).

comment number 7 by: polygeek

@Brian, In this case the ‘gods’ shouldn’t be indexed under the same URL. Try browsing the example again and you’ll notice that the URL changes every time the tab changes. That’s the Flex HistoryManager in action. So, hopefully, when this example gets indexed I’ll be able to do a Google search on this site for ‘athena’ and link directly to that specific tab.

Now, most Flex apps are not state aware and there are very few Flash apps that are state aware. In those cases all the content in that swf will be indexed as one url.

Hopefully more and more developers will start using the HistoryManager now that it really has an impact on SEO.

comment number 8 by: Brian

Yeah, I get the use of the HistoryManager, but that only changes the bookmark. Will Google really notice the change of the bookmark, and will Google care enough to actually index the pages based on the bookmarks? In the past, a URL with parameters defined a unique page (?article=gods) Now we are expecting that bookmarks define unique pages. (?article=gods#section=zeus) This does not work well for standard HTML pages, because the URL containing the bookmark is returning exactly the same text as the URL without the bookmark.

If it works that way for a HistoryManager-enabled Flex app, then it will be something new. That’s really my question. Can googlebot recognize a change to the bookmark, and will it treat that as a separate page? Adobe has said in http://www.adobe.com/devnet/flash/articles/flash_searchability.html that you can use a sitemap to define different entry points, but that’s not a complete answer.

comment number 9 by: polygeek

@Brian, Google recognizes changes in the URL. How else would they be able to return searches on this, or any other, blog?

Anyway, we’ll see when things start getting indexed exactly how it will work.

comment number 10 by: Brian

Because this URL is unique before the #? Do a search on site:polygeek.com and I think you’ll see that every URL that gets returned has nothing after the #. Whereas in a history managed Flex app, all the deep linking info comes after #. If Google does not store anything after the #, how will it ever deep link into a Flex app?

comment number 11 by: gualan

i wish you good luck with your expiriment…it is not so easy how it look likes

p.s. anybody knows that searchengines can understand colors on images, but i try to make an expiriment with searchengines text on images, they can't understand text on images…

comment number 12 by: polyGeek

@qualan, I haven't heard that any of the major search engines are doing image analysis. I think Google indexes images based on filename, alt-tag and the surrounding content. But someday I'm sure it will be done.

comment number 13 by: Henry Davis

Good luck.

comment number 14 by: SEO Packages

I think this is a very underlooked section of seo. People ususualy think about words and images when trying to rank and forget about file types such as pdfs and other files. Videos are becoming big too and I belvie search engines are trying to show more and more diverese media results.

comment number 15 by: Jimmy G. (Seo Blog Advice)

Does this really work? What are the results? Anyway thanks for sharing this idea and I wish you luck with this experiment.

comment number 16 by: Noah | WealthNet Partners

Hmm…pretty sure they've already been able to index swf files for some time now. Maybe what you mean by "truly indexable" is that they are now being spidered, crawled, and having their links followed? SWF files used to be indexed all the time, but now the contents beyond the name are actually being read and crawled for relevenacy and proper rankings.

As for image analysis, I believe that is something Google, specifically, is working on (but don't quote me on that one).

comment number 17 by: polygeek

@Noah, yes, I meant truly indexing dynamic content, which they now do.

comment number 18 by: web design + seo

I am really interested to see how this turns out. Mi coda su coda.. you crack me up!

   Welcome back (Change)

Leave a Reply

comment feed RSS   subscribe to this comment thread

Recent Posts

   



polyGeek.com

© Copyright 2008 polyGeek.com / Dan Florio, All Rights Reserved Except Where Explicitly Stated
Web Developement Blogs - Blog Catalog Blog Directory
M2 Websites
Local Directory for Los Angeles, CA

Better Tag Cloud