August 2012

How Much DITA and/or CMS Do You Need?

CIDMIconNewsletterTom Aldous, Adobe Systems, Inc.

Calculating ROI

In nearly 20 years of consulting, prior to my current role at Adobe, I worked with many companies, large and small, who had a compelling need for structured authoring. Over time, the preferred solution shifted from SGML to DocBook to DITA. A common pattern I observed is that it is easy for customers to become distracted by new technology, product features, “trends,” and overlook the fundamental needs that they need to address when selecting a solution.

Each client is different. Although there may be some basic workflow patterns that can be replicated from one customer solution to another, types and usage of content can vary greatly from one customer to another. A careful decision-making process is required for success.

There are a number of pitfalls that many clients fall into. A good consultant or vendor will help customers to steer clear of various “pot holes” along the way. This article covers some recommended practices as well as common pitfalls to avoid. At the end I share some methods you can use to calculate Return on Investment (ROI) when migrating to topic-based authoring.

Defining Constraints on Your Goals

While analyzing your legacy content, current authoring methods, and to whom the content needs to be delivered, limitations or constraints on your goals should be uncovered. It is a great temptation to try to define a DITA or CMS solution that “does everything” and anticipates any type of metadata that may emerge over the next 5 years. The greater the scope and complexity of your projected solution, the more difficult it will be to achieve a working pilot project or to even define a window of time when ROI can be achieved.

Vendors or consultants often need a “champion” or “advocate” within the customer task force who can assist with internally communicating reasonable goals to stakeholders who don’t work directly with the content involved.

Defining Your Needs

In top-down driven strategies, executives may sometimes want a consultant or vendor to tell them what they need. Any customer task force involved in the discovery phase defining a DITA authoring or CMS solution needs at least to identify its own pain points. The internal task force also needs to be able to determine (a) why the information is created, (b) how it will be used, (c) how it will be reused, (d) who will access or use the content, and (d) what the skill set of the content creation staff is (for instance, how complex can your authoring solution be?)

Any competent vendor or consultant can provide suggestions in these areas, but only the customer or solution end-user can accurately answer the questions listed here. Many times the needs are simpler than originally thought, which can lead to a simpler, more constrained solution, versus trying to “do everything.”

Estimating Possible Consulting Time

Almost any powerful DITA or CMS solution will require some consulting or assistance to get started. What you want to avoid is “open-ended” consulting or training with no defined end in sight. Here are some simple ways to avoid this:

  • Have the consultant perform a simple demo using some of your content (not a full pilot project; just make sure that you see your content in a demo).
  • Have the consultant demonstrate the steps required to convert your existing files into structured data. Determine how much of this process you can do yourself.
  • Have the consultant share any methods of customizing the authoring user interface to
    • temporarily hide complexity of structure and metadata
    • eliminate display of menus that may not be used by certain staff
    • know exactly what the user interface will look like when accessing reusable data from a CMS
  • Ensure that the consultant demonstrates the process required to customize or control formatting for various forms of output. If this process involves extensive creation of XSLT style sheets, there may be considerable time and consulting involved. Determine if you can leverage any of your existing formatting solutions.
  • Have the consultant demonstrate what would be involved to transform your content into various forms of delivery (PDF, ePubs, or WebHelp, for instance).

If you have your vendor or consultant demonstrate at least a partial solution with your content from this list, you will have some idea of how much outside assistance will be required during project implementation.

Planning a Realistic Pilot Project

Once you have defined your goals and needs (discovered during the discovery phase), work with your consultant or vendor to define a pilot project. A pilot is critical for several reasons. Although your full investment in your DITA or CMS investment may take some time to recoup, your upper level stakeholders will want to see something concrete within 6 weeks or less.

A good pilot project should include (a) conversion of content into structured data, (b) some authoring of fresh content, (c) a minimal demonstration of content reuse, with or without a CMS, and (d) content output to at least two of your required deliverable platforms (PDF and ePub for instance).

Example of a Constrained Customized DITA Solution for Training Materials

One of my last projects before coming on board at Adobe was for a major consulting firm that created custom authoring solutions for state and Federal government employees. The challenge was to assist a state government department that created training materials to migrate away from unstructured Word authoring to some form of DITA that would allow extensive reuse of content from topic-based authoring.

We designed a customized DITA solution that used “self-evident” element tag names, like “List of Materials” and “Material,” and so on. In addition, the structure rules (implemented in FrameMaker) constrained the document component choices presented to the author. We were able to customize workspaces and use special config files to eliminate menus that could have caused formatting mischief. Incidentally, the point here is not the product involved, but what is possible when there is a high degree of definition and a list of precise goals for DITA implementation.

Only five days of training were budgeted for a group of trainers who had never seen FrameMaker before; they had only used Word. None of the trainers had used DITA before. The customer was somewhat nervous about whether this would be enough support, so they contracted for 40 hours of Help Desk support; they only used 3 hours.

Minimal training and support were required because this project was unusually well-defined and constrained. The solution was intended to accomplish one thing: to produce consistently structured training materials with a high degree of reuse. In addition, the authoring solution was “self-documented” to a high degree because of “purpose-driven” element tags.

This project may seem a bit extreme in its simplicity and amount of constraint on goals, but it demonstrates what is possible when a discovery phase identifies precisely what the team needs to accomplish, and avoids the “Swiss army knife” syndrome of a “do everything” implementation.

Sensible Guidelines

I recently hosted an eSeminar with guest Joe Gollner of Gnostyx Research. Joe made several key points that are not commonly discussed:

  • Content Management is a supporting piece of an overall Content Solution Lifecycle. This lifecycle includes (a) Content Acquisition, (b) Content Strategy, (c) Content Evolution, and (d) Content Delivery.
    • Content Evolution is very important. You do not want to select a DITA or CMS solution that is so rigid that it requires major changes to architecture when new requirements (like ePubs delivery) arise.
  • The following rule of thumb will help enforce an agile approach to Content Management:
    • For every dollar spent on the solution, a second dollar must be spent on making content and process improvements.
    • For every two dollars invested, there must be four dollars returned.
  • Joe Gollner’s “Iron Laws of Content Management”:
    • Content Management must transcend any one CMS solution or it isn’t real Content Management.
    • Make sure that you fully understand your content and what you need to do with it before you commit to a CMS.
    • You can only understand your content by actually working with it (note: this can be particularly relevant where legacy content is concerned).
    • Tailor your investments in Content Management to what helps you be efficient.
    • If you’re not useful, you aren’t really managing anything.

Many of Gollner’s guidelines may appear to be common sense. So why don’t more people practice these principles? Getting back to my opening statements, many people become easily distracted by tantalizing technology and trends.

Calculating ROI for DITA Authoring

Although a massive ROI may not occur within the first year of a new project, there are several areas where a carefully selected DITA authoring solution can show proven savings:

  • Content localization
    • With greater content reuse, fewer words are translated.
    • A tremendous amount of post-translation corrective formatting can be eliminated.
    • Savings are multiplied by the number of target languages you publish to.
  • Single-source publishing
    • Your solution should require minimal development to create desired formatting or display for the many platforms you may need to publish to: PDF, WebHelp, ePubs, and so on
    • Changing content in one set of source files will correct all of your various outputs, from ePubs to PDF
  • Greater writer productivity
    • Your content creators should be able to access reusable data in a seamless fashion and assemble documents more quickly.
    • Reusable content (conrefs in DITA, for instance) will eliminate individual mistakes.
    • Writers who don’t have to deal with the complexity of manipulating “tagged” content can create content more quickly and comfortably.

This last point is particularly important and often overlooked. When evaluating a DITA or CMS solution, make sure that your internal task force includes content creators who will actually use the final product. It may be a natural instinct to pick the most technical engineer you can find to head such an effort. But keep the “human element” in mind; your migration to topic-based authoring or reuse of document chunks will only be as successful as your staff’s acceptance of the solution. Ultimately, only the author or content creator can judge whether a new authoring solution is going to work or not.

Final advice: be wary of anyone who steers you toward defining an internal task force that does not include authors and content creators.

Incidentally, I have created an online ROI calculator available at <> that will help you estimate cost savings using Adobe FrameMaker vs. some more traditional solutions that require extensive style sheet development. You can use this ROI calculator model to supply metrics for other product solutions as well. CIDMIconNewsletter


Tom Aldous

Adobe Systems, Inc.

Tom Aldous is currently responsible for worldwide Adobe Technical Communication Suite, FrameMaker, and RoboHelp Business Development and Evangelism. He manages the Adobe Technical Publication Business Development Managers and Product Evangelist. Prior to coming to Adobe he was Managing Partner of Integrated Technologies, Inc. for over two decades.