A webinar given jointly by MarkLogic, Aberdeen Group (research), and empolis summarized where the industry stands on the move toward offering users dynamic technical content -- dynamic in the sense of filtered for their needs and assembled into useful printables. The dimensions for filtering are the usual suspects: user skill/type, context (such as a specific product), and task type (setup vs. operation vs. troubleshooting). Aberdeen's research on current industry practices (albeit focused on the needs of manufacturers) showed they are being driven by these factors (most to least): speed to market, market segmentation, customizability, precision (need to cut irrelevant doc), and globalization (localization needs). Those organizations found to be "best in class" are pursuing these types of initiatives:

• Reduce sheer mass of content, for precise doc offering
• Use graphics to replace text wherever possible
• Single source content (70%), and structure per DITA (30%)
• Keep in-house project management of outsourced localization
• Create interactive and 3D graphics
  
• Integrate docs with product management data and live data sources, to pull/single-source relevant content
  

Despite the self-interested need to control spiraling localization costs, many other benefits ensue from these strategies. Much of the energy is going to solving the problem of precision (how to deliver the key 20 pages rather than force a download of 300 pages). The demonstration showed a Help interface that users could filter by product/area and then selectively add sought topics to a PDF for downloading/printing; it did this at the source, without an obvious mechanism for distributing this capability across customers' servers, for downstream customization. Also demonstrated was how filtered content was served up on the limited interface of a smartphone.

Key to this whole dream of dynamic delivery is, of course, structured authoring and metadata tagging: without structure down to a fairly granular level, no dynamic documentation filtering or assembly is possible. The first step for all tech writing groups, I think, is to implement a content management strategy that supports the possibility of these and other new delivery systems.