I have had a number of years to watch the digital asset management market develop since my first Phrasea project (an early digital asset management system that never won mainstream popularity despite many novel features) back in 1997, followed by MediaManager, Cumulus, and MediaBank implementations that same year. In recent years, development efforts in the digital asset management market have principally revolved around OEMing third-party products (ImageMagick, Adobe InDesign Server, TeleStream's FlipFactory, etc.) into the applications.

Digital asset management vendors have also been working to expose Web services and improve their APIs in response to a market that is hungry to integrate digital asset management systems into other systems (Web content management, customer relationship management, editorial workflow, e-learning platforms, and marketing resource management systems, for example).

But where are the Eureka! inventions in the marketplace? And will any of the vendors be able to keep up with the new demands that the market is placing upon them? 

I think there are several trends worth watching. Unfortunately, since most digital asset management system vendors are relatively small companies with limited resources, or smaller divisions of larger companies, no single supplier is positioned to capitalize on all of these innovations at one time.

Here's a short list of these trends:

Leveraging XMP 

Adobe's eXtensible Metadata Platform (XMP) is a specification that allows for embedding and therefore transporting metadata along with an asset for the purpose of interchange across systems. While many of the vendors have been paying lip service to the XMP specification in past years, we will see wider adoption in 2008 and 2009, mostly driven by customers' need to send metadata along with a file to different systems and external enterprises. This trend is being driven by a few needs in the market, for instance: 

• Media companies and marketing communications departments need to track digital rights more effectively and have those rights travel with assets
• As content producers seek to unlock the value of their digital assets through reuse as well as to present rich media from disparate sources, it becomes essential for semantic metadata to travel along with those assets

Ever since the XMP spec was first released in 2005, commentators predicted the standard was on the verge of taking off, but it never quite did...until perhaps last year. Within the past year, virtually all of the digital asset management implementations my company has executed have included the XMP spec to varying degrees. It might be that we're influencing that as a consulting firm, but more often than not the customer is driving.

So what's happening? An emerging, richer set of available services and standards is making XMP much more ubiquitous. For instance, increasingly digital asset management systems are not only reading the XMP packet of a file, but allowing the user to edit those values. Several of the products on the market allow certain XMP fields to be locked, some to be encrypted, some to be editable, and some to trigger scripted functions. There are a couple of vendors that are also allowing organizations to extend the XMP resource for custom metadata fields as well as enable XMP for digital assets that natively do not support the standard.

Version Cue

A little less than ten years ago, digital asset management vendors were embedding their functionality right into desktop creative tools such as Adobe Photoshop, QuarkXPress, and other popular applications. These plug-ins promised a great user experience since nobody would have to jump out to a separate digital asset management interface -- likely a thick client -- to manipulate assets (checkout/checkin, update, etc). I was very excited about this approach back then and equally disappointed when it was subsequently abandoned by the vendors.

I think what happened was that individual desktop plug-ins were just too expensive to maintain and develop. The desktop applications were constantly going into next point releases that would require the digital asset management vendors to upgrade the modules for each application. The situation got even more complicated when customers requested customizations or enhancements of these modules. And plug-ins are notoriously difficult for enterprise customers to support as well.

Enter the Adobe Creative Suite and Version Cue. In its earliest incarnation, Version Cue was more of a prototype, without a usable API, that promised a workflow tool embedded right into the tools that creative teams used -- namely Adobe Photoshop, InDesign, InCopy, Illustrator, and Acrobat. In its most recent version (CS3), Version Cue (along with Adobe Bridge technology) offers a way for users to collaborate with check-out/check-in and version control. Various digital asset management vendors are currently developing solutions that are integrated with Version Cue and we should start to see more case studies published later in 2008.

We may also begin to see third-party independent software vendors (ISVs) tapping Version Cue's APIs to build extensions for non-Adobe applications. My concern for them would be that Adobe has changed the Version Cue code base significantly several times over its history and may not be done doing so, and I don't see the digital asset management system vendors pursuing that course either unless forced to do so at a specific client implementation. 

However, for companies that need to develop workflows into their Microsoft Office applications, there is another platform to contend with from our friends in Redmond...

The Ubiquitous SharePoint 

Microsoft reports that SharePoint has become a $1 Billion business with 100 million licenses sold. There are a number of use cases for SharePoint, but the one that relates to the discussion above is the ability to check out Microsoft Office documents and apply metadata within an access controlled environment that offers repository services like version control. Sounds an awful lot like digital asset management, except without other common features of digital asset management (compression, conversion, renditioning, video transcoding and keyframing, and so on).

Within the past 3 to 6 months we have been getting more requests for integrating digital asset management systems with SharePoint. My expectation is that several digital asset management vendors will begin to offer integrations of their products with SharePoint. Integration with SharePoint will enable companies to extend the workflow capabilities of their digital asset management system, as well as to further integrate digital asset management into the technical architecture of an organization's back office. 

I have seen a recent implementation where a digital asset management system was integrated with Microsoft's Connected Service Framework, a service aggregation framework that connects directly to Microsoft offerings like SharePoint and Exchange Server. 

Rich Internet Application Interfaces

At some point around 2002 or 2003, as the Web application tools began to mature and browser technology stabilized, digital asset management vendors saw an opportunity to get away from (or at least minimize) end user dependence on desktop or "thick" client software. From that date until the present, we have seen digital asset management system vendors strive to achieve feature parity with the thick clients of yesteryear. They have done so with mixed results, using Java applets, AJAX controls, or in some cases niche products such as WebObjects.

What we will begin to see in 2008 is that digital asset management vendors will invest in building Rich Internet Applications with technologies such as Adobe AIR and Adobe Flex to create desktop applications. These kinds of products allow vendors to build a single desktop client version of their end user applications that can run on either Macintosh or Windows, as well as connect over the Web.

Think Single Source of Truth Repository

Digital asset management systems have historically run as islands unto themselves, but that isn't good enough for the modern business. There are numerous cases where companies invest millions on their digital asset management system, millions more on their Web content management system, then still more on their workflow, collaboration, and other business systems. Then, they have to spend more on the integration of these different systems just to throw content over the wall to each other.

What our customers are asking for, at least, is that their digital asset management repository, which should hold the single source of truth for any piece of intellectual property, should be easily interconnected with other systems. Recently several of the vendors in the market have responded by beginning to integrate their digital asset management systems with Web content management systems, workflow managementsystems, and the like.

Going one step further, we see a short list of enterprise content management (ECM) vendors making the necessary investments to do away with the separate stores behind their digital asset management systems in favor of one repository for all content. These vendor investments will simplify the expensive integration burden that customers face when trying to connect business systems with their digital asset management solutions.

We will look to continued vendor announcements in 2008 that are consistent with digital asset management as an easier-to-connect-to repository for content and business systems, as well as comprehensive architectural integrations of digital asset management with ECM platform solutions.

Unlocking Article Content -- the role of the XML Server and Text Mining

Increasingly companies are looking to create textual content once then publish to any medium. Publishers, for instance, have grappled with article content being effectively "locked" in creative files as opposed to storing content in a central repository and instancing it in print, Web, and so forth. This costs organizations work effort and lost time since staff need to copy content out of one file format into another.

The XML specification has become the de facto standard for storing content in a media-independent format. XML server technologies have emerged as powerful repositories for XML content. These XML servers are easily integrated with text mining engine (TME) software. The benefit to this approach is that once a brand's taxonomy is defined, the TME can index the content against search terms that are meaningful to that brand. The results include improved producer or consumer search experience, easier mashup, and other serendipitous semantic linkages of content, and amplified results on Google and other search engines.

While digital asset management systems have traditionally been great for managing creative assets, they have fallen short in the way in which they handle XML or article content. Digital asset management systems essentially treat text files as binary objects without much visibility into the content. It is true that digital asset management systems have integrated third-party, full-text indexing tools to harvest content out of certain textual file formats. In general, however, digital asset management products do not offer functionality such as natural language processing or entity extraction from article content types -- particularly those assets that are in XML format.

In 2008 we will see clear trends where digital asset management vendors will partner with or perhaps even acquire suppliers of XML content repositories and text mining engines to harvest metadata from the article content objects as well as to make reuse of those objects a more rapid or self-selecting process. Consistent with the notion in the previous section that digital asset management should not be an island unto itself, we are seeing publishers look to digital asset management as a resource to index, store, and transform media-independent article content bundled along with other content types like images and videos. Since storing and transforming content is a familiar functional capability of digital asset management systems, it is a short step to extend digital asset management principles to the concept of article content types.

Where is Open Source?

There is one development, however, that is not happening: the development of a robust, commercially-supported, open-source digital asset management solution. Thus far the open source digital asset management projects like Fedora, DSpace and ResourceSpacedon't pose significant threats to the key players in the digital asset management marketplace. 

This is not the case in other content technology areas, where you can find thousands of developers working ostensibly for free to create custom applications that work with platforms like Facebook and open source Web content management platform projects like Drupal and Plone. Since many of the open source and open-platform projects have thousands of people swarming to develop the different pieces to the enterprise content management puzzle, it seems likely that sooner or later, the digital asset management piece to that puzzle -- including the vision painted here -- is going to be met by that approach as well.

In fact, we may see a current digital asset management system vendor take a page out of the playbooks of companies that are delivering commercially available open source projects. I am sure that some clever marketers could find a way to make at least as much money if not more on a services+support approach than what most of the digital asset management system vendors or digital asset management product divisions of ECM companies seem to be making in today's marketplace ($5MM - $50MM/year last I checked, which is relatively small).

Conclusions

None of these developments, individually, constitute a Eureka moment. Together they suggest a digital asset management industry working hard to get out of its creative niche and address the wider needs of the enterprise. This is not at all a new story, though.

What is new, from my vantage point: no single digital asset management system vendor has the resources to address all of these varied needs in a timely fashion. In the end, it may take the harnessed fervor of the open source community to bring all these threads and more together in the marketplace over the next 24 to 36 months.

(Written in 2008 by DPCI CEO Joe Bachana)