From the Director
Getting a Return on Investment: Progress with DITA implementations
DITA Europe 2010 was an outstanding success —in part because of several presentations about successful DITA implementations. Among the exciting presentations were those by Keith Schengili-Roberts of AMD and Vikram Nanwani of ITT Fluid Technologies.
Keith Schengili-Roberts reported on his fourth year of a mature DITA installation. From 2003-2006, Keith reports that they were using unstructured FrameMaker to produce their information. Their localization costs were very high with difficult quality assurance required for code pages. They could not keep up with product releases. Writers did not collaborate, with very little reuse of content, and the quality of their work was inconsistent.
Since moving to DITA and the Ixiasoft content management system, they have seen
- Productivity increases of 2.3 to 3 times what they were originally
- Fewer writers than they started with but higher quality work
- Localization costs less than half of what they were before the move to DITA
A key factor in measuring increases in productivity is to know what activities writers are engaged in, how much time they take, and how valuable they are for the customer and the product. Keith estimated that writers spent 50% of their time formatting content in their previous process. Moving to DITA virtually eliminated this activity.
In addition to using DITA and a CMS, Keith’s team also emphasized minimizing content, reusing content among document types, and discovering what customers really want from the information.
The CMS made all the activities faster. In the first year, documents produced increased from 5-8 per quarter to 14-16 per quarter, and the average time to produce the documents decreased from 70 to 80 days to 24 to 38 days. Output was doubled with the same headcount and an increase in document types. Productivity has continued to improve, with 23 documents produced per quarter in 2009.
Keith compares the work of an equivalent organization with 9 writers and similar subject matter. Using a process similar to Keith’s old process, this team produces roughly 40 documents per writer per year. Keith’s team of 4 writers produces roughly 90 documents per writer per year, making them 2.3 times more productive. Their two localization coordinators handled 432 documents in 2009. Not only did content reuse reduce localization costs, but also desktop publishing charges were eliminated.
The CMS more than paid for itself through the reduction in localization costs but Keith’s cost savings did not stop there. He has seen increases in the number of topics developed by the writers and more 100% matches for translation.
I hope that Keith will agree to develop his full presentation in an article for Best Practices newsletter soon, so that you can understand in detail how he tracks productivity. The details I’ve presented here come from his slide presentation at the DITA Europe conference.
Vikram Nanwani directs the Global Enterprise Content Management (GECM) organization at ITT Fluid Technology. At DITA Europe, he gave us a summary of the progress that his organization has made in their four years of DITA implementation. When ITT began its DITA implementation, the installation, operations, and maintenance manuals produced throughout the company, in many different divisions, varied widely. Translating the manuals was quickly becoming cost-prohibitive. Differences in content, writing style, and presentation not only made information reuse impossible, it also hindered the ability of ITT to encourage cross selling among its multiple brands.
The ITT vision, established in the GECM business case to senior management, called for “a single source of ITT Fluid Technology information available globally in reusable form, governed by common processes, tools, and controls for multiple publication outputs to multiple channels.” Their mantra became “create once, reuse anywhere.”
After a successful pilot project, the GECM leadership has had to continue selling the project to management, including new business acquisitions and new product lines. As a result, the team has built a program that includes
- Regular updates on actual costs, timelines, and future running costs
- Improved clarity of communicated numbers
- Measured “real” current costs rather than models
- Savings and soft benefits communicated clearly
- Demonstrations of how reuse reduces migration and translation costs of subsequent products
As a result of these analyses, Vikram has demonstrated that ITT experiences on-going savings of $1.5 million each year. Initial costs were $1 million below the original predictions in the business plan.
However, moving ahead is never easy. The executives forget the original business plan and continue to look for evidence of cost savings. Vikram noted that every organization seeking to implement the DITA standard and manage content more effectively must
- Spell out the problem and the need
- Build a vision
- Communicate continuously at all levels of the organization
- Develop real numbers
- Recognize that each project phase requires a new approach and re-selling
- Understand that the people in management change and forget so you must keep justifying the work
But, in spite of all the challenges—they made it!
DITA acceptance in Europe has lagged acceptance in North America, in part because many information-development organizations have not taken the time to study cases like those at AMD and ITT Fluid Technology. By studying the business cases now available, with three to four years of experience, organizations have the data to point to and the methods required to sell new projects. Leadership in these organizations must itself first recognize the value of topic-oriented, XML-based structured authoring. Without this recognition, the focus required to support a DITA implementation will not occur.
I see great potential for DITA in Europe, especially considering the number of companies in the machine industry, where the benefits of reuse and lower translation costs are obvious. However, it takes internal leadership from managers who know their present costs and can influence executives by building strong business cases. Any support that CIDM members can provide or information that I can make available is here for the asking.