History of single-sourcing Doc and Training in InfoDev

Just wrote this up for a presentation proposal for the Austin ASTD conference, so I thought I'd share it for the record:

For a decade now, our small Information Development group has worked toward the goal of single-sourcing content between user guides and technical training materials. Without content reuse, we found that huge disparities developed between these outputs: procedures would be updated and expanded for training that would never make it back to the documentation, and information would actually conflict, leaving the student to sort it out. Separate doc/training source generated double the work and a maintenance nightmare, and we were charged to solve it – affordably.

Our first generation of the solution leveraged the tools already in place to support single-sourcing of printed guides and product Help files: Microsoft Word and DocToHelp. Our solution for single-sourcing training was to blend the deliverables and to present a shared “Learning and Using…” guide as both user reference and training guide, the latter being supplemented with standalone labs, slides, and demos. We merged the learning content (objectives, exercises, self-quizzes, scripted walk-throughs) into the conventional documentation content. By necessity, information was structured to privilege the needs of instructors, and reusable learning objects didn’t exist below the granularity of section/chapter. Yet our Word styles were consistently tagged and significant rewriting already occurred to support non-linear Help topic access (removal of metadiscourse (“As we just discussed,…”) and sequence cues), so we were well on our way to having well-structured and modularized content – necessary for us to take the next step.

Our second generation of the solution leverages a content management system, AuthorIT, into which we migrated all of our existing Word content. AuthorIT is both an authoring and a publishing environment that manages content objects in a database, all for the cost of a desktop publishing package (an essential point, as we couldn’t afford enterprise CMS). Topics (procedures, overviews, definitions) are linked to templates that drive their behavior in the outputs, such as starting on a new page or in a new window; books (hierarchical structures of topics and other publishing objects) have templates that control not only the page layout but also output customization, through macros. Step one, we put the training content type topics onto special templates to suppress them from the online Help output. Step two, we organized reusable learning modules into shared book objects, so that new courses and organizations can be quickly assembled without incurring a maintenance hit (no worries about topics being added/deleted/moved in the user guide, since it shares the same book object). Step three, we implemented a version control system (SourceSafe) to organize and web-publish all remaining training deliverables that aren’t generated from AuthorIT. Finally, we solved the release target conflicts (documentation work focuses on the future release but training builds new outputs for the current one): the documentation group works out of the production SQL Server database that targets the upcoming release, while training works out of the snapshot database that froze documentation content at the current release level.

Today, we are researching how to create instructor slideshows and other interactive media to be shared and published through AuthorIT. We will also divorce the page layout of the user guides from that of the coursebooks and remove the learning content from the user guides altogether. We will implement world-wide on-demand coursebook printing. We will make editable Word outputs available on the intranet for instructors to customize at will, while locked PDF versions (all produced through automation) will be managed separately for printing/sale. We are also looking to expand user/task analysis and architect ways to single-source, say, a terse procedure for admins with the graphic-heavy, feedback-intensive version for staff. AuthorIT lets us clever our way to amazing results, and the efficiencies we gain from single-sourcing create the time to do it.