How difficult is it to implement a content management system integrated with a translation management system and using a structured authoring environment (DITA)? CaridianBCT can tell you exactly what it takes to do this and to do it well. In Part I (http://www.dclab.com/dita_global_local.asp), Jennifer Linton of CaridianBCT identified the various steps taken by CaridianBCT to create this type of environment while hopefully reducing costs. In Part II, she focuses on the overall process identification methodology and its eight phases, and documents both the technical and business processes undertaken to successfully launch the new CMS/TMS system, called GEM (Globalization and English Management System). Part II features include what CaridianBCT learned from previous efforts and describes the various approaches and pilot phase that were adopted to support the new GEM system in future efforts.

Part II

Identifying future concerns

As we move forward into our new environment we have taken a lot of precautions to identify risks and encourage a positive change management strategy, but we do still have some concerns and issues that we have to consider moving forward. Listed here are some of the future concerns that we anticipate.

• Existing localization service providers (LSPs) may resist the process change because they do not approve of our internal translated content management. They do not like the cost decrease in desktop publishing associated with our internal automatic publishing capabilities. They may add more to the overall project management costs to compensate, and they may start charging more flat fees because we are eliminating the cost of translating already translated content.

We learned that it is important to qualify the LSPs to ensure they understand our translation technology. We also implemented an in-country review process to promote quality assurance and ensure regulatory standards are met in different countries.

• In-country reviewers may not be able to perform reviews at a topic level. We hope the reviewers can learn to review at a more granular level, but they may resist the change because they feel more comfortable reviewing everything in context.
• Our stylesheets will require iterative development cycles to support multiple languages.

Our Content Management System and Translation Management system may have some pitfalls. Although we feel we purchased tools to improve our storage management and workflow in the new environment, we still need to anticipate enhancement requests to ensure problems do not occur during live projects. We also have to continuously and extensively test the CMS integration with the TMS to make sure we do not lose any data passed between the systems. Additionally, we need to always be aware that new tools are hard for everyone to use. Authors, reviewers, translators, and in-country reviewers all need training, and usually they need training more than once.

Calculating cost savings

Although we anticipated some concerns, we have actually identified quite a few more benefits to implementing the system than drawbacks. In our pilot project we estimated nearly $100,000 savings for two deliverables in 9 languages (approx. $5,500 per language per deliverable). We are saving this money primarily with automating desktop publishing and decreasing translation costs due to leveraging translation memory and internally managing previously-translated content. We also see many benefits from reusing content at the topic level where we can reference the same content in different hierarchies to produce different deliverables. As a result, we only have to create the content once, review it once, translate it once, and then can easily produce it over and over again thereby greatly improving not only the efficiency, but the consistency as well. The tools also provide a mechanism for the authors to have more control over the content ownership. However, the tools and technology make it slightly more difficult for Subject Matter Experts (SMEs) and reviewers to make minor changes because the new systems also impose greater control over the change process enabling CaridianBCT administrators to now also track the significance and impact of any, even minor, changes using these systems.

"Flexibility and portability are also large benefits as a result of our new environment. CaridianBCT may now reach a broader customer base more readily."

Flexibility and portability are also large benefits as a result of our new environment. CaridianBCT may now reach a broader customer base more readily. We can now create different products for different audiences using the same authoring system regardless of customer locations, formats (PDF, HTML), and even different languages. Flexibility also includes the ability to set parameters for the filtering of content in many different ways. Since almost identical content is created and translated and is only filtered by what is required in house, CaridianBCT may easily publish specific product versions for different audiences using the same content just setting a different filter type for the various versions required by each distinct audience. This process uses less time, less resources, and therefore, less money. Just as the content may be shared among different customer types, the new environment also allows us to easily share content among different departments in the Company. Now, Marketing, Training, Web Development and other functional areas use the same content and methods to create information, improving consistency company-wide across all deliverables to all customers.

Remembering lessons learned

During the new environment rollout, we of course added up the list of lessons learned. Change management is usually the obvious lesson. The transition from document- to topic-based authoring does not just affect the technical documentation department. Other departments involved in the overall process included the reviewing team, compliance and regulatory department and the translation team as well as many other areas within the company. Each affected department also needs to learn about topic-based authoring, so make sure you have a topic-based authoring evangelist to help spread the word and conduct training. Also, until people start working directly with the tools, there is only so much that can be done with the concepts and theories. Therefore, you will need to have some content available to begin working with early on in the launch.

The pilot project development process itself triggered many of the lessons learned. Firstly, it helps to conduct a pilot project where you can test changing the content. We couldn't change the content too drastically to follow the DITA methodology without increasing the workload to create a change list for regulatory requirements from our customers. Additionally, instead of writing content to fit the DITA model and our information architecture, we had to change the DITA model somewhat to accommodate our content. We didn't necessarily create a customized DITA, but in some cases we just didn't use DITA best practices for authoring. The plan is to go back and revise the DITA for the next release. Next, we also may have set our goals pretty high for just one pilot project. It is important to not try and test all use cases with one pilot. However, if you take this approach, it may extend the pilot project and decrease the opportunity to demonstrate quick wins, maintain a sense of urgency, and maintain deadlines. In our pilot project, we were able to test many of our case studies, but there are still many pieces to test and we have to create more pilot projects to test the remaining case studies due to timeline creep.

"...in a regulated company, it is also extremely important in the new environment to know the regulations that drive your documentation."

Remember, too that in a regulated company, it is also extremely important in the new environment to know the regulations that drive your documentation. It helps to partner with the internal Regulatory department to understand how regulations are interpreted at your company. The regulations and interpretation may vary depending on the organization. Finally, the last business lesson learned is that it is important to hire dedicated resources with the right skills. Make sure you have someone with a technical background to help roll out the technical pieces of the puzzle.

Moving forward, we are excited to expand our GEM environment to additional departments including training, quality assurance management, marketing, service, and more. We hope to decrease the bottleneck caused when ECOing (ECO = Engineering Change Orders) our deliverables by ECOing our topics at a more granular level allowing us to create our deliverables in less time. Finally, we hope to identify additional methods for improving our content delivery to our users such as HTML/Web, PDA/Mobile devices, embedded help, CD, and more. There is so much more opportunity available as a result of embracing this new environment.

DCLNews Editorial
December 2008