a thousand furnished rooms

Just because these arguments can be mapped across a few decades--back to the online information services of the 1970s, through the first blush of CD-ROM in the 1980s, and right up to the Internet and Web and enterprise portals of today--doesn't take away from the force of these convictions. On the other hand, after so long a time this argument has been made--and as variously applied as it has been--there's a certain impulse to say, Been there, done that.

In fact, despite the presence of new digital content delivery platforms in the form of e-book readers, there is little new coming out of such conferences about e-books that goes much further than offering--ironically--an electronic analog of the print book. Never mind some of the new wrinkles being brought to bear in the digital publishing scene, of which digital rights management (DRM) has been thrust to the fore, right along with (and in the case of XrML, in combination with) XML-based content tagging and management systems.

As important as DRM and standards-based content management are to rational, efficient, and cost-effective document and information serving, and yes, even if that document is a book, there remains one challenge that often still comes up short: getting users of digital document systems the exactly right content these users need at the exactly right time these users need it. While it is a great idea to get any content seeker the content he or she seeks, most of the real action of managing digital information is taking place within companies that have a real ROI interest to motivate good content handling, and among these businesses' partners and value chain participants.

Giving Content in Context

For enterprises wishing to benefit from the creation and management of content portals, the challenge is clear. Systems that manage content without managing the context fail.

Searching for content is a frustratingly difficult and easily overwhelming exercise. This is true even as search engines are increasingly bolstered by technologies and processes to help make them more effective--spiders, meta-data, standardized taxonomies, and human editorial intervention. The problem of course is an ever-greater avalanche of data. The projections for simply and effectively finding content are dire, and hardly a case of Chicken Little; for example, where there were less than 200,000 web sites in 1995, there are, a half-decade later, 22 million, and these numbers don't include most intranet sites that are closed off to Web indexing efforts by firewalls.

The Web's promise (among others) is to improve communication both within and outside the enterprise. To succeed, however, customizing the content delivered to employees, partners, and customers becomes important.

Getting Personal about Content

Personalization requires enterprises to have the means to capture information--the term "profiles" is typically used--about the information users. These profiles need to be useful in directing specific content to those profiled, which means that an enterprise also needs to know about its own content and, if used, third party content.

There are many elements that can be used to deliver content in context. These include:

� Registering content meta-data for Web and enterprise-wide search engines
� Implementing effective search engines (e.g., relevancy)
� Collecting and managing profiles of site users (e.g., personalization engines)
� Creating and maintaining taxonomies of content (e.g., subject classifications)
� Identifying communities of interest (e.g., portals)
� Enabling pass-along content delivery (e.g., superdistribution using DRM)
� Sending email content offers/links to profiled users

Some companies rely on powerful search engines that possess tools such as relevancy ranking, natural language query, built-in thesauruses, contextual hit results, and other improvements to the electronic searching.. Other companies simply rely of self-selection of its content users, where the assumption--as in the case of many enterprise and vertical portals--that the focus of the site carries enough implied context. The more effective solutions, of course, are those that use as many contextual content delivery strategies as possible.

The more robust, detailed, and accurate the meta-data, the easier it is to find content in huge content bases and return find hits and serve the content itself. If such search effectiveness is tied to personalization profiles that track a content user's interests and requirements and content delivery mechanisms, content delivered in context becomes powerful indeed.

But for enterprises today, perhaps the biggest benefit is gained by mastering how enterprise content can be served into specific contexts within the business process and partner chains, to deliver more on the promise of automation. Look for such management of content (which could be called "syndication") to play a growing role in tying the information of the enterprise to the many different parts of the enterprise's business actions.

(My thanks to David Guenette, who collaborated with me on an earlier version of this article.)

Bill Trippe
btrippe@nmpub.com