Process Improvement in a Vacuum

	Process Improvement

by Michael Wood

November 19, 2001

Situation: The company's business rules are buried in its legacy systems, and no one understands them anymore. Given that, how do you conduct a process improvement project and achieve breakthrough outcomes?

This was the challenge that was recently faced by a process improvement team. What follows is how they overcame what appeared to be impossible obstacles.

Normal process improvement projects center around the facilitation of management to develop quantitative objectives and the facilitation of knowledge workers on how end-to-end work process could be improved to achieve those objectives. The business systems that support these activities are typically centered around work routing and maintenance of customer orders, products, etc. The business and processing rules are well understood, even if only by a few key staff. In this project, the knowledge workers knew their industry and how to interface with the systems, but no one understood how the systems produced the results they did. The architects and designers were long gone from the company. The algorithms were buried deep in the code, and documentation of the data structure and programs was nonexistent.

To overcome these shortcomings, the project team had to improvise and pursue alternative tasks not normally part of a process improvement project.

The project began like most. The management objectives were quantified; the workflows were mapped and tested. A gap analysis revealed areas where substantial labor savings could be gained with a new system in place. And then the wall was hit. The systems were "Black Boxes." Data went in, data came out, and the transformation rules were unknown.

The team's first approach was to work with internal IT staff to understand the database design and supporting data dictionaries. Sounded like a good idea, but the team hit another wall. There were no file structures--just text strings--and no dictionaries to decode the strings. The next step was to perform code reviews and look at data labels assigned to data strings being used. Another dead end. The programs referred to a piece of data in terms of its position in the string and not by its data name. Therefore when the team would ask for the meaning of a piece of data, the answer would be something like "4^th record, 3^rd position." The team found it comical and yet extremely frustrating.

The final approach was to reverse-engineer every output into its normalized data structure. This produced a database design that could be used to understand the gaps between inputs and outputs. Typically the database design is not an output of the first phase of a process improvement project. However, in a vacuum of knowledge, it becomes the center of the universe for conceptualizing new systems.

This effort became a discovery process. The focus of facilitation efforts were purely on "what should be." It took many work sessions with production staff to free their minds of the constraints they were living by. But about three weeks into the project, the breakthroughs came. This happened when the team presented a conceptual framework of how the production of products could be structured, which was a dramatic departure from current methods. The result was a recasting of the project from process improvement to organizational reengineering.

The team learned a valuable lesson from this project. It learned that the knowledge of an organization exists in many forms and many places. While people hold the keys to how things should be, they do not always have a clear understanding of how things work today. Certainly they understand the procedures they follow. However, in a world of seamlessly integrated technology, where business rules are obscured from view, process improvement efforts often must extend into very technical areas where road maps are not easily found.

Today process improvement teams must have the skills and tools to deal with organizational change, people facilitation, industrial engineering and information technology. They must be able to discover knowledge at all levels of the organization. Most of all the teams must be able to shift gears rapidly and adapt their methodology to the situation. Creative problem solving is perhaps the most important skill the team must have. In the above case the team pursued unconventional techniques to achieve breakthrough results.

Staying open-minded and flexible while maintaining focus and direction is the message here.

What obstacles have you encountered in your Process Improvement projects? What have you done to overcome those obstacles? What were the results? Let us know. Share the knowledge.