Phased DITA Implementation Increases Manageability… and Your Odds of Success

By Naveh Greenberg, Director, US Defense Development, DCL

Implementing a new technology in an organization requires planning and complex choreography. Choosing to move to the Darwin Information Typing Architecture (DITA) standard for authoring and maintenance often involves more than one technology, and always a significant shift in how content is viewed and created in the organization. If you’re planning a DITA implementation project, you already know it’s not for the faint of heart. But following some common sense best practices can make it more manageable, and increase your chances for success.

We recently conducted in-depth interviews with teams at 12 companies that went through a DITA implementation to pinpoint best practices that would aid other organizations considering the move. And, while they all considered their implementation projects successful, they unanimously agreed it always takes far longer than anticipated, and requires a solidly planned phased approach to make it work.

Milestones are a given for any technology project, and selecting the content management system (CMS) and authoring tools are key milestones in a DITA implementation. We found that half of the companies we studied selected their CMS at the beginning of the project, while the remaining conducted program pilots for converting content, and extensive testing to understand their specific data requirements prior to selecting tools. The other half followed after various program pilots. Two of the companies changed their decision on the CMS during the testing phase, underscoring the importance of a phased approach to the effort.

Your earliest phases of the implementation should result in solid understanding of your current content—structure, formats, currency, and quality—and the processes your creation and production teams currently use to publish content. You’ll want to identify the potential for content reuse, cited as one of the major business drivers behind DITA implementations conducted by our survey respondents. These companies also indicated that, although most were not actively translating content currently, ease of translation is a significant business driver.

Will your DITA tools work “out of the box?” The answer for most of our companies was a qualified yes. Understanding business requirements led several to look at additional specializations to allow them to take even better advantage of the benefits of structured content, appropriately tagged. So understanding your own business requirements and developing a specification to address them becomes another key phase in implementation.

Because DITA requires a shift from traditional content authoring under a print paradigm to structured authoring based on smaller components of content, training and change management become critical throughout all the phases of a DITA implementation, but especially early on. All of our respondents agreed that training the content authors proved critical to their ability to ramp up production. At DCL we follow a proven project engineering methodology that aligns with best practices for DITA implementation, phases that make the project manageable:

1. Thorough data analysis

2. Business requirements and conversion specifications

3. A “proof of concept” phase using a small subset of content manually prepared

4. Software customization

5. Proof of production sample

6. Conversion prioritization and scheduling

7. Production ramp-up

   
   

We work with clients to create checkpoints with thorough testing and relevant metrics for each of these phases to ensure that you know what to expect and how to best leverage a fully implemented DITA environment.