As covered on Slashdot and elsewhere, Adobe annouced today that it has provided technology to search giants Google and Yahoo! that will enable them to crawl and index flash (swf) files. It seems that the 'crawlable content' within flash movies are limited to the text elements and links and doesn't include images and videos.
From a broader perspective it is worth considering what, if any, impact this might have on the landscape of search results and of search marketing. It's difficult to believe that even with the ability to crack open flash movies and consume the text content within them that search engines are going to have an easy go at making sense of what's inside and appropriately assigning their relevance on search engine results for a few reasons below.
Deep Linking
While it has long been technically possible to allow for deep linking on flash-based website it is certainly not common or in all cases simple or straightforward. The 'named anchors' feature rolled out with Flash MX that intended to natively support this capability was a mess of browser support and is next to useless. Using some of these techniques each page has a different URL but contains the exact some flash movie, using javascript to load specific frames. It's doubtful that search engines are going to be able to take this into account and recognize 'different pages' within the movie. This obviously has a significant impact on a flash site's ability to rank well across a number of keywords.
Lack of Semantic Structure
The flash platform and the resulting movies lack the defined structure inherent in HTML making it more difficult for search engines to assign the weight and value of one text element over another. While you can extract all the text elements into a big word soup, it's likely difficult to analyze the resulting data for much more than frequency and because of this you can run into the problem of sites exploiting this.
Opportunities for Black Hat Trickery
Google and other search engines have become pretty sophisticated in sniffing out things like keyword stuffing, keyword hiding, and so on. It seems that it would be extraordinarily difficult to uncover similar techniques within flash movies. Since it seems likely that simple word frequency would be, at least initially, the best and/or only way to determine relevancy this could be problematic. If flash sites ranked on par with non-flash sites then there would be a lot of incentive and opportunity to game the search results via flash.
Conclusion
While it's great that Adobe and search engines are working together on opening up all the content currently hiding within flash movies, there are a lot of technical problems that would need to be overcome to meaningfully and appropriate index flash. In light of these issues I hope that, while still indexed, websites done primarily in flash do not de facto have the same weight in search results as sites that are not.