DITA Learning and Training: A strategic overview
In recent years, DITA adoptions have been on the rise with an increasing number in varied fields outside of software. What’s next? With more companies already realizing the value of structured, single-sourced content in product documentation, leading edge companies are increasingly looking at training. Some have either set up custom XML or repurposed the DITA 1.1 concept, task, and reference model for learning and training deliverables. Others are experimenting with features of DITA 1.2 already publically available in beta.
The upcoming release of DITA 1.2 includes the new Learning and Training Content Specialization. A mature beta version is already available in version 1.5 of the DITA Open Toolkit. The specialization provides a way of packaging and structuring DITA that makes sense for training, both instructor-led and computer based. DITA Learning and Training is a beginning point, not an end point, a kind of toolkit for solving specific problems of scale, delivery, and maintenance, which many companies encounter ever more frequently as they grow.
At the CMS/DITA Conference in Santa Clara, California this year, I demonstrated the latest work from the OASIS DITA Learning and Training Subcommittee. As DITA Learning and Training is a toolkit and general approach, I presented an argument for real world training samples. Realistic samples provide an excellent starting point for organizations to evaluate DITA Learning and Training as a possible training solution, for tools vendors to explore potential feature support, and for anyone wanting an example of what the new specialization is designed to do (and not do).
With training deliverables covering processes well outside of technical documentation, DITA Learning and Training has the potential to reach a far wider audience of DITA adopters. In this article, I discuss the need for well-defined best practices in a learning and training implementation. I entertain a documentation maturity model that includes both product documentation and training content. Finally, I touch on what skill sets to look for in the new world of combined and coordinated documentation and training groups.
What is DITA Learning and Training?
To put it simply, DITA Learning and Training is a superb example of DITA specialization at work. Though the specialization builds on the standard concept-task-reference model of DITA, it packages topic content in a way that is optimal for training. It proposes a best practices model with a standardized structure for topics, lesson progression, and content reuse. Though beyond the scope of this article, the specialization also supports SCORM, a standardized approach for packaging online training.
Who would be Interested in DITA Learning and Training?
DITA Learning and Training solves problems that not all organizations may be experiencing (though, if they are growing rapidly, and especially if they are expanding into international markets, they will probably experience them soon enough). Organizations will be interested in DITA for training for many of the same reasons as DITA for documentation.
Many companies want to separate content from presentation for simultaneous, automated outputs and deliverables (such as in HTML and PDF), to support reuse and repackaging of modular content, and to lower translation costs. In general, companies with a need for increased automation and for flexible packaging options will be drawn to DITA Learning and Training.
Though many organizations are initially drawn to DITA for efficiency and cost reasons, companies like IBM and AMD have documented dramatic increases in customer satisfaction after their move to DITA, indicating that modularity and reuse improves the quality of content. DITA is not just about cutting costs.
Less duplication of content means less maintenance and fewer mistakes reaching customers (XML-based content also provides a rich environment for error-reporting against semantic content, a feature that companies are only beginning to exploit). Catching errors before they go out to the public improves quality. Plus, goal-oriented topics inherently focus on the real world needs of users. Giving audiences exactly what they need and when they need it improves the user experience.
DITA Learning and Training is, after all, DITA, so the same advantages that apply to DITA in general apply to Learning and Training. DITA Learning and Training provides a robust architecture for reuse, repackaging modular content for varied distributions, while also supporting simultaneous outputs in various formats. Growing companies often experience the increasingly expensive burden of content maintenance based on legacy systems that are better suited to smaller (non-localized) deliverables. DITA-based information systems, on the other hand, are in their element when scale is involved.
It is true that DITA Learning and Training, like DITA generally, requires the initially complex task of setting up an automated build environment. Though, once created, the build can rapidly crank out various, simultaneous deliverables from a common source automatically, such as Instructor-led (ILT), computer based training (CBT), web sites, PDF books, and so on.
Furthermore, once content is available in XML, now the preferred choice of many translation agencies, the same build system can also generate translated and localized versions of the same training. Given the increased pressure for companies to support simultaneous translations and localized deliverables, DITA provides a competitive advantage to realize revenue within the same quarter of the core release.
Though there is usually increased setup cost and initial complexity over a legacy desktop publishing-type solution, the cost savings downstream, especially for larger organizations with numerous products and translations, can be astronomical, easily eclipsing the cost of an entire writing or training department. DITA Learning and Training is a flexible, sophisticated toolbox solution for high-scale operations with global reach.
In addition to its cost, efficiency, translation, scale, and quality advantages, DITA Learning and Training is increasingly of interest to companies that wish to share content and resources between documentation and training groups. As tight schedules and increased product deliveries continue to squeeze content groups, the need for efficiency is greater now than ever. DITA Learning and Training promises increased leverage of common deliverables and content in a common format of XML. Though a common content infrastructure might seem like a holy grail for many stressed documentation and training groups, there are some critical challenges to surmount on our potential hero’s journey.
Adopting DITA Learning and Training
I’ve been involved with numerous DITA and structured XML conversions over the years. Some of the same problems that plague DITA adoptions in documentation trouble training initiatives. Odds are that existing content is not highly structured or consistent, especially if stored in a vast legacy of desktop publishing formats.
Many companies will initially try to fit old content into the new DITA bottle. However, Learning and Training, like DITA in general, is founded from the ground up on notions of consistency, modularity, reuse, and multiple channels of distribution, in other words, on a consistent set of best practices. Rather than start from a clean slate and determine what is the optimal structure based on a standardized model of content delivery, many training organizations are saddled with huge legacies, often in Word files and PowerPoint slides, that may or may not follow a consistent strategy.
I have also found that typical training deliverables are inherently more complex than most documentation sets, and thus subject to more inconsistency in approach (if you think in DITA terms, most training provides more content types than online documentation). Add company acquisitions to the mix, and it is likely that training approaches and assumptions of what works best vary, sometimes passionately, from group to group. In the worst cases, training is a free-for-all of whatever happens to fit into the desktop publishing bucket. Just because something fits in a set of slides does not make it training, let alone an expression of best practices.
DITA Learning and Training conforms to a validated, consistent structure. Unlike most desktop publishing solutions, it does not automatically lend itself to whatever shape of content happens to exist. It is especially important when considering a DITA implementation to identify the optimal structure, at least for new projects, which can proceed unimpeded by the inconsistencies of yesteryear. Supporting legacy projects, where it makes sense to convert them and bring them forward, will require additional creativity.
Fortunately, DITA Learning and Training provides a high degree of flexibility for organizing files independent of presentation. For instance, the Learning and Training content can form a type of shell that pulls content from other base topics for concepts, tasks, and reference, enforcing new structure on old content. Training maps and content references, key tools for reuse, can then be used strategically once legacy content is broken down into base topic types.
Meanwhile, new training can proceed with an optimum structure based on the specialized training topics. However, you may want to consider storing all new base training content in standard DITA topics as well if reuse between documentation and training is critical. What can be shared where between documentation and training is a critical question that needs answering prior to a new implementation (more on this later).
It’s Just Programming (or Scripting)
One of the challenges of DITA Learning and Training is that the specialization does not look like a traditional training tool (like Word or PowerPoint). It is after all, a specification that any number of commercial or open source solutions can use. While we can expect an increasing number of commercial tools to leverage the Learning and Training specialization, early adopters and information architects today must contend with XML in its rawest form, likely using tools that are better known in documentation circles already using XML.
While training authors can readily use a more familiar pseudo WYSIWYG environment (any XML or DITA-compliant editor will do), there is no getting around that orchestrating a powerful build environment for training requires expertise in XML (for content), XSLT (for general transformation), probably XSL:FO (for PDFs), and likely Ant (for builds). While an open source solution like the DITA Open Toolkit might provide a solid, do-it-yourself foundation, additional work is needed to tweak XML processing to generate the specific results that you want. Commercial solutions may hide some of these technical requirements in the near future, but early adopters today need to roll up their sleeves and work the open source toolbox.
Given the immense power and flexibility that the toolbox approach offers to companies, both today and tomorrow, this is not a bad thing. Granted, there is likely an upfront cost in moving to open source scripting solutions from former automation environments based on VBScript, for example. On the other hand, with the continuing spread of DITA, there is a strong argument that the scripting skills of the future for both documentation and training will be based on open source solutions like XSLT, XSL:FO, and Ant.
Even if companies ultimately purchase the newest whiz-bang commercial offering that does most everything out of the box using DITA Learning and Training, being able to get under the hood and optimize XML for special purposes, now and in the future, is a powerful, competitive capacity. If I were building a training organization today, I would ensure that infrastructure skills for XML, XSLT, XSL:FO, and Ant were readily available.
The Need for Real World Samples
Because DITA Learning and Training is not an authoring tool, but a methodology, it is especially important to understand how the specialization works in the real world. Out of the box, DITA Learning and Training provides sample content that is useful for verifying XML processing (for instance, using the DITA Open Toolkit or a new commercial offering). You can also get a good example of a best practice implementation of training and a standard flow of specialized content types. As a basic sample, it is not intended to show off all that is possible. Nor does it include fancy formatting, though nothing would stop you from incorporating whatever sophisticated presentation is required for your company.
For evangelizing DITA Learning and Training within companies, it is best to develop your own sample content with company branding, hopefully based on company best practices already compatible with DITA. If current practices are not compatible, it may be time to review how DITA Learning and Training can add value by promoting the latest industry best practices. Prior to actual implementation, it is best to develop a full-blown prototype with all of the desired features in DITA Learning and Training. Once the prototype is complete, it can function as a template for the rest of the organization.
More samples based on real training content would help early adopters visualize what is possible. As part of the OASIS DITA Learning and Training Subcommittee and the DITA Adoption Technical committee, I am heading up an initiative to develop and gather real world training samples to give better ideas of how to structure content in DITA XML. While the sky is the limit, with embedded videos and simulations, I expect to start simple. If you are interested in joining this initiative, contact me at email@example.com.
A DITA-based Maturity Model for Training and Documentation
As training and documentation become more rigorously structured, conforming to the latest best practices for modularity and reuse, you may find that the two formerly disparate worlds of the technical writer and the trainer look increasingly alike. Documentation has become increasingly goal-oriented, focusing on practical tasks that meet the real world needs of its audiences.
In the mission critical ERP world, the training I developed has always been goal oriented. While it is possible to fill tomes of content about what a complex application or suite can potentially do, if users can’t go back to their workstations and perform their jobs today, the implementation is a failure. The proven solution to mission critical training environments is a highly structured approach based on goals. I was thoroughly familiar with structured approaches and minimalism years before I encountered DITA.
If documentation and training are looking ever more alike, sharing content seems like a no-brainer, but there are pitfalls. The issue concerning consistency of approach comes up once again with sharing documentation and training. In many companies, writers for documentation and training work in different departments with little interaction. The documentation and training models may have significant overlap, performing or duplicating similar work for the same customer. You might think that inherent duplication is a good case for re-use, but in many cases, the additional duplication serves no real function when you consider that the same user typically has access to both training and documentation.
I have seen training manuals with content that would have been better served as reference material in the help system that users can access any time while at their workstations, after training. I have seen content in documentation that would be better served as full-blown training courses, either online or instructor-led.
In some cases, highly specialized products require training that all customers will receive with purchase. Instead of trying to duplicate the function of training, the help system in this case is better served as a support for the user’s learning after training.
Different companies, products, and services will have an inherent information mode spanning documentation and training that is best for their purposes, but not always fully realized. Ideally, training and documentation information models reinforce each other, not duplicate functions.
When I worked on SAP implementation teams at the beginning of my career, I was fortunate enough to work on both documentation and training at the same time. Duplicating work was not an option. I did not have time. The documentation and training models I used, rather than focusing on the duplication of re-use, eliminated gaps and reinforced each other with their respective strengths.
Online help was more generic, comprehensive, and universal, whereas the training focused on real-world scenarios conforming to a specific business use case. The training did not attempt to cover every conceivable product option and feature. Though the online help contained a comprehensive description of critical tasks, the training fleshed out a critical subset of tasks under a real world scenario the audience could immediately relate to, ignoring product features and options that did not support the particular scenario.
Users can always explore the help later to see all the other features that the product can support. Once they have a mental model, it is much easier to put minimalist documentation included in Help (or within the UI) into context (getting started type information in documentation can serve a similar function in products that do not have formal training).
It might seem that tasks can be easily shared between documentation and training. After all, the training uses the same tasks covered in the documentation. Discussing DITA training models with some early adoptions, I’ve encountered a number of people whose initial enthusiasm for sharing tasks was quickly blunted. They found that the training tasks were often too specific to share with the more generic help system.
On the other hand, if training content becomes too generic for the sake of reuse, customers have a more difficult time developing a mental model from which to apply their knowledge in the real world. If documentation and help try to cover every possible real world business scenario, the content becomes increasingly verbose, difficult to navigate, and ultimately impossible to consume. In the end, only some generic concepts could be readily shared along with some reference material.
In an ideal world, the documentation and training works together with little overlap, each doing the job that they do best and not each other’s jobs. A maturity model supporting training must take the documentation model into account. With a comprehensive model, combined documentation and training departments can work optimally on related content.
DITA and the Future of Cross-Functional Teams
Understanding an information model that optimizes documentation and training also speaks to the skill sets of each group. Companies that wish to combine the two groups under DITA should not assume that a good product writer is a good training writer or that documentation deliverables can be easily repositioned to fulfill the role of training.
While the tasks and goals may be the same, good training follows through with a rigorous learning model of demonstrations, practices, and evaluations, all reinforcing the learning goal. I think of these additional training dimensions as the value-add on top of core documentation (of course, product documentation can also be specialized to include additional content types, but most companies stick with the more generic concept-task-reference model).
Can product writers learn to write training deliverables? Of course. But it should not be presumed that product writers already know the best practices of a training development model. They will need training. Conversely, training writers must scale back on content types when writing lean product content. There is no reason why one person cannot perform both roles, fully coordinating documentation and training content types, though product and department size, as well as release schedules, may well dictate specialized, separate roles by necessity.
In the years ahead, I expect to see more documentation and training groups combined to share knowledge and approaches whether the same writers provide both documentation and training deliverables or not. Even in the case of specialized roles in larger companies, coordinating the information model will allow better communication between product and training writers, avoiding unnecessary overlap and maintenance. The end result is leaner documentation and training that is easier for users to digest and use in the real world.
The Future of DITA Learning and Training
The future of DITA Learning and Training looks bright. With international markets opening across the globe and DITA penetrating fields far beyond its origins in computer software and hardware, Learning and Training addresses pressing, expensive problems that are not going away.
At the same time, there are significant challenges that impact how we think about training in particular and content in general. Throwing legacy content into new packaging does not necessarily fulfill the potentials of newer XML-based content systems. Those companies that are most ready to transition or even jettison older practices, while promoting the increased use of best practices, will experience the most competitive advantage.
DITA Learning and Training isn’t about adopting a new authoring tool. It is a global approach to content, both training and (inherently) documentation. Companies need to rethink how they work with both documentation and training, outside of the silos, so that they work optimally together, avoid unnecessary duplication, and best serve the customer.
For the latest status of the specialization and for available samples, see the OASIS DITA Learning and Training Subcommittee webpage:
Troy Klukewich is an information architect for Oracle Corporation with extensive experience in enterprise software, structured documentation, and mission critical training supporting live implementations. Prior to Oracle, he was the senior documentation manager for Borland Developer Tools. He has also worked for PeopleSoft and on numerous SAP implementation projects for PricewaterhouseCoopers. He can be reached at firstname.lastname@example.org.