Imagine what it's like to manage an information repository encompassing over 100 titles, and several hundred megabytes of data, and captured in formats as diverse as PDF, hardcopy printed materials, computer software delivered on CD-ROM, and web services. Just ask Jennifer Jung, Content Technology Manager at J. J. Keller & Associates, Inc. Jennifer was the Project Manager for transforming not only content, but also the systems and processes managing the vast array of content and content types produced by J. J. Keller.

From Humble Roots to Publishing Behemoth

J. J. Keller has been providing support compliance tools and services to heavily regulated industries which now include transportation, manufacturing, environmental, distribution, chemical processing, construction, food safety, human resources, workplace, mining, and more since 1953, when Jack Keller, and the newly formed J. J. Keller & Associates, first opened as a one-man consulting shop. From this humble beginning, an information behemoth was born that now, 55 years later, has grown into a world-class publishing, printing, and service organization that's grown to over 1,200 associates and serves more than 300,000 customers throughout the United States and Canada.

Today J. J. Keller offers more than 6,000 products and services to help customers meet their safety needs and comply with regulations enforced by DOT, OSHA, DOL, EPA, FDA, and 300 state agencies. In order to manage the complexities associated with providing critical information in these areas, J. J. Keller's experienced and knowledgeable editorial department is staffed with more than 25 experienced editors, all experts in regulatory compliance and safety issues.

"...offers more than 6,000 products and services to help customers meet their safety needs and comply with regulations enforced by DOT, OSHA, DOL EPA, FDA, and 300 state agencies"

Its product line contains a vast array of different types of content including bills of lading, construction regulations, state hazmat information, general OSHA requirements, forklift standards, crane safety, DOT safety, environmental regulations, tanker inspections, and much, much more. Data is also created and offered in several different formats.

Making the case for Single Source and XML Conversion:
Proof of Concept Redux

Back in January, 2002, Jennifer Jung, Content Technology Manager for J.J. Keller, and several colleagues from multiple departments were summoned into the boardroom to talk about forming a Total Quality Management (TQM) Task force. The task force would investigate how to establish a single-source environment that would enable J. J. Keller to introduce efficiencies in how the content was produced and maintained, from initial creation to when the content actually left the building.

Although the single-source concept had been tossed around before, this time there was an even more compelling rationale for moving forward. Jennifer notes that "Because many of the existing content support tools had reached their end of life for support by vendors, and since J. J. Keller's plan was to create a truly media neutral content repository, it became an even higher priority to determine how Keller and its content could be more efficient as new areas were addressed including web services, new channels, customers, aggregator opportunities, and even getting products to market faster. We sensed that single sourcing was the answer, but still needed to prove this concept to executive management and even ourselves."

The ability to use the same content in multiple ways was the obvious benefit to a single-source approach. Jennifer elaborates further, "In the body of the thirty-eight print materials alone, all of the titles share some pieces in varying degrees, with some only duplicating a section or a chapter while others might contain close to 65% overlap. Imagine if there were no duplication and each individual use application needed might be easily pulled from the one and only one, original source document". As Jennifer added, "We wanted to have the ability to repurpose the same content for different formats, and use the content repeatedly-perhaps as a book, or a loose-leaf, or even a small pocket guide covering workplace safety, or forklift operation. Everyday it seemed there was another new opportunity to build a different product from existing content that resided in several different places. We wanted a system that would enable the content to be iterated over and over again, in as many formats or platforms as possible, but residing in one, easily accessible master repository. In the existing environment, the tools used to compose pages involved editors writing in one format, while another area, using a different tool set, would need to reconvert the content in order to reuse it. "

"we wanted to have the ability to repurpose the same content for different formats, and use the content repeatedly - perhaps as a book, or a loose-leaf or even a small pocket guide covering workplace safety, or forklift operation."

Great ideas, but the approval to move forward and actually implement this type of system still needed to be granted. To do this, a pilot program was authorized to evaluate the concept and validate the efficacy of a single-source approach. All the then-current work processes were meticulously documented to assess the initial state of J. J. Keller's systems; what was being used and where, and what processes surrounded the creation of that content. The pilot ran from May, 2003 through September, 2003. Portions of three publications were selected as the test bed. Each department area had an equal say in order to ensure buy-in for the system that was eventually selected and launched. "The main concept was indeed validated", said Jennifer, "and it was determined that adopting a uniform format, specifically XML, was the way to go. More importantly, we would be able to produce what was needed for multiple publications from one set of common tags used around the content. It also became clear that a main imperative would be to create content that was tool independent or technology agnostic, and that could be utilized across multiple platforms and distribution models."

Obtaining Awareness and Buy-in Across the Editorial Organization

"It also became clear that a main imperative would be to create content that was tool independent or technology agnostic, and that could be utilized across multiple platforms and distribution models."

Jennifer was appointed Project Manager with the task of identifying and implementing a solution that would enable J.J. Keller to transform its data management process from both a production and a maintenance perspective. The envisioned end game would provide tools that would then maintain the life cycle for the content produced in a more efficient and effective manner.

With content growing exponentially, especially since regulations change constantly and materials need to be updated; this was a daunting task to say the least. Since many of the J. J. Keller product lines feature the same regulations, albeit translated to address a specific industry, content duplication across titles occurred frequently.

"Our goal was to create an environment where Total Quality Management combined with single-sourcing could be applied to the content production process, and thereby provide J. J. Keller with a more streamlined and efficient publishing process. Although we believed there was significant duplication of content across multiple sources, how much and where it was located was impossible to determine using the old systems. Even internal editorial teams were often unaware that similar content that might be reusable existed in other sources. The obvious human productivity gain presented another key benefit for the entire company."

"Our goal was to create an environment where Total Quality Management combined with single-sourcing could be applied to the content production process and thereby provide J. J. Keller with a more streamlined and efficient publishing process."

We asked Jennifer about the enormity of this project, the challenges and obstacles that she had to overcome initially, and the even larger challenge that any new system would require J. J. Keller editors to change the way they work. Jennifer continues, “Obtaining buy-in from the editorial teams was one of the toughest challenges especially in the initial phase of the project. People don’t like change, and we were suggesting a change to the way they had been working for years. Therefore, editorial teams were directly involved in the overall assessment of the new authoring tools and processes. “

Goodbye PDF, Hello XML: Data Conversion Project is Given the Green Light

The pilot was indeed a hit, and the go ahead was granted to move forward with the complete project.

The conversion project was broken into four phases.

• Vendor selection; tools and resources for conversion
• Conversion plan
• Testing - quality of output was critical
• Identify duplicate content across sources

Other questions also needed to be answered including:

• Would an off-the-shelf CMS really work at J. J. Keller?
• Where could J. J. Keller obtain the biggest ROI?
• Could a uniform DTD be created and used?
• What would the new workflow look like if it was proven successful?

The content and DTD used for the pilot were then sent to each potential conversion vendor for testing. Vendors were measured based primarily on quality, and their flexibility in managing changes. "Data conversion tends to be the most challenging part of a project" indicates Jennifer, "especially when moving loosely formatted legacy content into a new highly-structured format. Although this may be accomplished with programming or even a combination of automation and human intervention, it is still one of the more difficult aspects of a conversion project. The legacy data itself may be difficult to work with and contain errors. We wanted to assure that output was not impacted by junk in, junk out."

As it turned out, the conversion was actually the least of J. J. Keller's worries. Data Conversion Laboratory, the vendor ultimately selected to do the work, "really shined" said Jennifer, "DCL just kept getting better and better with each conversion, and as we soon found out, DCL was also able to help us plan the conversion and focus on where the greatest Return on Investment (ROI) would be early on by identifying that content with the most opportunity and overlap."

"DCL was also able to help us plan the conversion and focus on where the greatest Return on Investment (ROI) would be early on by identifying that content with the most opportunity and overlap"

The DTD and subsequent structured XML allowed J.J.Keller to compare content at a designated "chunk" level to determine duplicate sources. It also enabled the ability to link or single-source at a chunk level. Business rules with system automation provide the mechanisms necessary to perform maintenance and police content usage to make sure that Keller's single-source goals are sustained long-term.

The End of the Beginning: Was it Worth it?

Where is J. J. Keller today? Extremely pleased with the XML repository, Jennifer adds, "It's still a work in progress. Some issues still exist with batch composition - configuring this is tough especially for print-ready content. We needed a lot of infrastructure to support this, but have learned a lot and adopted new practices, processes and even a few new roles.

A dedicated team is in place to support the system and stay current on upgrades, etc, but this is no different than what was done before, just the technology and the processes used have changed. No new editorial head count has been added. Right now Keller is attempting to marry two repositories together, an older repository for web services and software that is still HTML based, and the newer repository that outputs XML. A bridge solution currently translates the XML output into HTML. At some point, the vision is for one, fully accessible repository available to everyone at J. J. Keller who needs it.

Moreover, the ability to address new channels, channels that J. J. Keller typically didn't market to, will provide even more exposure and brand extension for the company. Jennifer concludes, "J. J. Keller may now use and reuse content in more ways than ever. We can offer different versions of content for different platforms, chunks of data for syndicators and aggregators, and create updates outside of the regular publication schedule-which is very important especially with government regulations that may be rescinded at any time. All in all, the entire project has already proven its worth.

"the ability to address new channels, channels that J. J. Keller typically didn't market to, will provide even more exposure and brand extension for the company."

A core component of any single source effort is the system's ability to identify similar content across what may be very disparate source materials. Coming in the next issue of DCLnews, Jennifer Jung will return with Part II of "How J. J. Keller Struck Gold; Using DCL's Harmonizer Tool." Here Jennifer Jung will elaborate on how J. J. Keller used DCL's Harmonizer tool to eliminate duplication across its publications, thereby providing an additional source of ROI for its content and conversion processes, perhaps even more importantly, how J. J. Keller used Harmonizer to help plan the order of their conversation to give them the biggest bang of their re-usability buck.

BE INSPIRED AND HAVE FUN!