From the Director
It’s easy to claim that investing in new tools will increase productivity. Tools vendors promise that your writers will work faster, producing more content in less time, if only you will author in XML, automate your workflow, store content in a repository, or develop a single-source architecture that incorporates a reuse model. Because most of us work inside high-tech companies that produce new software applications, we are enmeshed in the technology-centric belief system that says automation will save the world.
Unfortunately, although structured authoring and content-management applications can provide useful tools that make information easier to manage and maintain, they rarely do so without initial costs. Without developing more effective processes to plan, manage, create, and deliver information, you will have difficulty justifying a return on your initial investment.
In this issue of Best Practices, we focus on process changes. We ask managers to scrutinize current practices for opportunities to remove cost. We ask the hard questions:
- What is the cost of each step in the current process?
- What value is produced by that step?
- Is the value produced worth the cost?
If we find steps in our process that produce little value to the customer, we must consider finding less expensive ways to do them or eliminate them entirely.
Information developers and technical publications managers have long argued for the importance of what Vesa Purho calls “the hygiene factors.” We are concerned that if our readers find spelling, grammar, or other basic errors in our documents, they will come to distrust the reliability of all the information. As a result, we devote a significant, measurable fraction of our resources to hygiene. We certainly rely on basic tools, such as spelling checkers, to catch the most egregious errors, but we know that tools alone are insufficient without a process that includes human copy checking.
To achieve consistency in format across media, publications organizations have invested, rather heavily it appears, in standardized markup language technologies, initially SGML and more recently XML. In most instances, the goal seems to be focused on producing outputs in many media without the need for manual conversions. From a single SGML or XML source, we can design outputs to HTML, PDF, print, and others as needed, all without any manipulation of the source files. By avoiding tools that use proprietary format codes, we reduce the cost of production. Many organizations report production time expenditures decreasing from days to hours by eliminating tedious handwork and multiple final production error checks.
In such solutions, we reduce the process steps or eliminate them completely.
Organizations look for ways to increase the return on their investments in markup languages by single sourcing. By this, we mean using chunks of content in more than one information product deliverable. A paragraph originating in one document becomes available for use in other documents. Voila! Cost reductions.
At the same time, however, organizations pursuing productivity increases through single sourcing are frequently disappointed. We’re achieving only 1 percent reuse, we hear from managers. How can we do better? The problem, of course, is not the tools but the lack of a new process for planning and creating information.
Yes, you say. Let’s develop modular content. Rather than embedding information in whole documents and waiting hopefully for reuse to happen, perhaps there is a better way, one that requires a process change. Can we create modules of content and apply them effectively to different information products? What sort of process change is implied? Writers want to know how they are supposed to create disembodied modules in the first place.
The process changes needed to move to modular development are complex. We need to learn to think both about the design of parts and the design of the whole-often simultaneously. Design is like that-you have to
consider everything at the same time. Compare, for a moment, designing information to designing a mosaic. Mosaics are constructed of hundreds and thousands of tiny chips of colored stone. Each chip is pretty nondescript looked at as an individual, but its shape and color are determined by its ultimate place in the whole.
If we look at the mosaic with all the chips in place, we see something quite miraculous-a complete, often stunning picture. The whole becomes much more than the sum of the parts.
Can we use the same chip set to develop another stunning picture? Yes, of course, for the most part. Some chips can be used in more than one picture; some chips must be designed uniquely for the true message to emerge.
Should we just sit around the stone pile and create chips, then? Not a very interesting job certainly and probably not very productive. We don’t know if the chips we create will have any use. Rather, we need to plan the pictures and design the chips at the same time, keeping in mind that some of the chips will work well in other pictures of comparable intent.
A new process? Of course-a process that includes someone who knows how to create useful pictures and someone who is an expert at carving out chips. Working together from the outset will make both more productive and the end product of reasonable cost. The chip maker will be focused, as will the developer of the individual content modules. The picture designer will define the raw materials needed to build the best picture for the viewer. In like manner, the collections of print and PDF books, linked Web sites, and help systems will come together quickly and effectively.
Do we need the support of well-designed tools? Yes, but only after we understand what we are trying to accomplish and have a process in place that we can demonstrate without the tools. Once the process is worked out, we’re ready for all the help we can get.