The art of retrieval
The proliferation represents a double-edged sword. For the first time, corporations have the capability to store and manage branding, marketing logos, promotional and training videos, and other digital content in-house — instead of paying outside agencies to store those assets in geographically dispersed, hard-to-access locations. That creates new revenue opportunities by making it easier to repurpose those assets for distribution on Web sites, kiosks, cell phones, and other channels. Those benefits can only be realized, however, if the right assets can be searched and retrieved by the right people — a big “if” when talking about tens of thousands of images and other sorts of unstructured data that consume terabytes or even petabytes of space.
That’s where DAM (digital asset management) comes in. DAM uses a methodical taxonomy of metadata — and in some cases voice recognition and optical character recognition — to make it easy, say, to locate an obscure video from two years ago or the exact slides used in an old marketing campaign.
But DAM goes well beyond that. Once the province of media, entertainment, and marketing professionals, a host of new industry players are embracing DAM and using it in rather innovative ways. Just a few of the newer applications for DAM systems: Call centers that use phoneme recognition to identify the cause of a large spike in complaints (think “battery failure” in a computer manufacturer’s support center), pharmaceutical companies that organize and distribute digital content to the FDA for approval of new drugs, and casinos that index vast amounts of surveillance video. In order to create a repository for digital media, the ultimate challenge is to create a structure where it didn’t exist before.
“There’s still that saying that a picture is worth a thousand words,” says Stouffer Eagan, CEO of Autonomy, a provider of DAM systems that have the capability to read license plate numbers captured in video, analyze phonemes within audio, and automatically generate smart tags based on concepts in an underlying search result. “If it has in it what you need informationally, it’s probably worth 10,000 words.”
Autonomy’s ability to deduce patterns and relationships was particularly attractive to managers at the NewsMarket, an aggregator of video for journalists in print and broadcast. The product, which Autonomy markets under the name IDOL, reads pre-inserted metadata to help reporters find content related to SUVs, for instance. But it can also recognize that some of the images emphasize fuel efficiency or were featured at the New York Auto Show. Says Shoba Purushothaman, president and CEO of the NewsMarket: “IDOL tags things with stuff we didn’t think of.” The New York-based company also uses the product to make suggestions based on an individual customer’s browsing habits to help find video suited to their precise needs.
A quest for collaboration
Because of its highly unstructured nature, rich media tends to be a mishmash of interrelated video, still images, and PowerPoint presentations. It can be particularly hard to manage when it’s created and edited by a large team located in multiple locations throughout the world. Furthermore, digital content frequently needs to be stored in a variety of formats, from broadcast-grade video to compressed files suitable for streaming. That means DAM systems must also be able to handle transformation, check in, access privileges, and workflow processes.