I’ve had some conversations with vendors recently that revolve around the way the people are implementing DITA. The question was basically “Are people jumping in with both feet, specializing DITA to fit their content? Is there a phased development?”
We are seeing a lot of phased development that begins with a move to DITA out-of-the-box. Companies who translate their content can gain a very quick return on investment by moving their content quickly to XML and component-based development. They save money by converting content to topics and other components and translating only what changed. The also save money by automating the publishing process — reducing desktop publishing costs that can be incredibly high particularly if you translate. So, they can get a quick return by moving their content to DITA as-is, without specialization.
But most feel that is only a starting point. DITA, without specialization, does not always provide the authoring control that they feel they need nor can it always sufficiently describe some of their key information. They can point to the money that they’ve save on translation as the funding they need to go the next step. The next step typically involves properly analyzing and modeling their content, determining the DITA specializations (content/metadata) that they need to ensure they can produce and manage their content effectively.