by Barry Schaeffer
reprinted from Newspapers and Technology magazine.
In newspaper circles, firms like The Bureau of National Affairs, Inc,.in Washington, D. C. wouldn't get much attention as a model for online editions and techniques. BNA, traditionally a publisher of targeted information and news content for businesses, professions and government agencies, doesn't take advertising and doesn't cover the local sports teams unless they're involved in litigation or mergers. Instead, BNA publishes nearly 200 information products from daily, on-line only newsletters to multi-volume tomes for attorneys, accountants, etc. So why should the newspaper industry look to such a firm for tips on how to proceed down the path toward an online presence? The answer lies in the fact that, for the differences between them, BNA and its newspaper cousins have many things in common and confront many of the same challenges in serving their markets.
Perhaps most important, as much for bona fides as substance,
BNA supports a daily news gathering and analysis function
that would do justice to any newsroom or wire service. With
several hundred reporters supported by remote
correspondents, the firm collects breaking news in many of
the same areas covered by the primary news press (the White
House and Congressional press corps, for instance, always
have BNA reporters among their ranks.) The major difference
between this news organization and its newspaper
counterparts is the absence of layout. BNA publishes its
paper editions on letter size paper with neither color nor
complex layout considerations. Not having to worry about
layout has been catalyst for much of what BNA has done in
the content delivery area. Drill into this situation and you find a company that has
over 50 years experience in how to design, collect and
deliver content. Some elements of that experience hold
lessons for newspapers moving toward an online presence.
I'll capsule some of them here.
Content counts: Starting in the late 70's, BNA acquired a number of technologies to help it produce paper products that would compete well in the growing market for targeted information. Back then, computer systems traded data format for performance, using highly proprietary formats to support their functional goals. As the market moved towards electronic delivery and immediate access, software vendors struggled to keep up. The result was often marginal products, longer than reasonable schedules and high cost. For a firm like BNA, known for delivering targeted information to a demanding market, this became a serious and growing problem. No amount of customization or manual intervention could fully close the gap between what the readers wanted and what the systems could produce. Finally, in the mid80s, BNA adopted a new "system independent" content protocol called "SGML." The firm had realized that, in the end, it was creating content that had potentially broad and lasting value if only it could be managed effectively. No proprietary system, however elegant in its chosen function, could or would do that. The lesson in this evolution is that the end product is often not the entire story and the true value in many endeavors lies in the underlying content. A subtext, inherent in the time it takes to make the conversion to rich content, is that one can't afford to wait until the market demands change to start the change process.
From Creator to Consumer should be a smooth flow: It's almost an article of faith in the information industry that if you want a certain output, the best way to get it is to create the raw material in a form as close to the final product as you can get. The further away from their deliverable state data are captured, the more intervention and cost is required to make it ready for the user. As clear as this seems, the publishing industry seems to learn it all over again with each step in the evolution of information products. The most common scenario, one that BNA partially replicated in the 1980s, is to design and build the slick output and management portions of the system, then if there's money and time left over, address the creation side of the process. The result is skilled authors using word processors (or worse) to capture complex data intended for inclusion in sophisticated information products. The result is almost always frustrated authors, production people angry at having to clean up errors the authors couldn't catch, rework editors putting codes and other items in the data. BNA faced this situation through the 80s and into the early 90s, using a variety of low-level tools to generate input in SGML and suffering for it. In the early 90s, the company decided, at the highest level, that it should spend its efforts on designing a flow that allowed writers to create content to develop and produce the kind of information products the market was demanding. Newspapers are facing the same challenge today, hazily understanding that subscribers want targeted information on tighter schedules and with a web of links to related information around it. Most newspapers have found out how difficult this is when the raw material must be reworked before it can be used. What many haven't realized is that their dilemma comes from upstream in content capture, and there they must go to deal with it. Just deciding to use XML, for example, doesn't do much good if you have to rework your writer's content to get it. Worse yet, if what the writers create simply cannot be sufficiently enhanced to produce the desired output, you have figuratively nailed one foot to the floor and will likely go nowhere.
Take the long view of things: There may be some quick hits with the online world but producing a "sticky" and profitable electronic newspaper isn't one of them. When the decision is made to build a truly web-based information environment, a newspaper is embarking on a process that cannot be completed properly in a few months or with a couple of new system buys. BNA, to the credit of its management in the early 90s, realized this and set up a schedule and resources equal to what they saw as a long and demanding task. The firm did a number of discrete things, like setting up a special group within the company to drive the new project. Although not quite as disconnected as Lockheed's fabled "skunk works," this group could undertake efforts outside the normal flow of production, develop and mature concepts without day-to-day scrutiny or politics, and make its case in the best possible light. Production people were transferred into the design group on a rotating basis, ensuring that the results fit within the company's business goals and resource limitations. Most important, the firm gave itself the time to work out the issues and build something that its users could really handle. The result has been one of the most successful projects of its type in history.
Today, BNA has converted most of its many products and staff to the new system, developed new products based on the near automatic recombination of it new data resources and publishes much of its output nightly on the web and several electronic media and has created a staff, both editorial and technical, that understands both the how and why of electronic media. Newspapers would do well to understand how BNA did it and what parts of that experience have value for the newspaper industry. Layout or not, BNA has blazed an important trail and the lessons are there for the taking.