Dan Madison is a principal in Value Creation Partners. He facilitates process improvement using lean, six sigma, reengineering, and continuous improvement techniques. Dan is the author of Process Mapping, Process Improvement, and Process Management.
Dan Madison has studied what a business process should look like for fifteen years. He studied what major corporations did in process improvement that made them successful and distilled his findings into design principles that anyone can use. He has come up with 38 design principles that apply to all business processes.
His first example was from the permit and inspection department for the city of San Jose, California. Before the redesign, it had taken more than a month to get their permits approved. There was no money available from the city for process improvement. Madison helped the city develop a new permit form and process using four “lenses,” or pictures of how the processes were done and how the results were judged. These were:
-
- Frustration (as experienced by the people in the process)
- Quality
- Time
- Cost
Frustration level is very important because Madison says that it has a high correlation to quality and there is usually immediate buy-in from people within the process.
After finding out the frustration level, it is important to find out why. In the permit process, the main problem was incomplete information on the permit forms, so the process was amended to ensure that all the information that was needed was in place from the beginning.
The developers needed to have a single point of contact with the city, in order to find out the information they needed. The city also began to cluster the permits so that similar projects had their own process. This meant that a developer who wanted to build a high rise had a different process and form than a homeowner who wanted to put in a swimming pool. It was found by studying the processes that the no-permit-needed people and the simple permits were found to take half the permit load for the city. With minor changes and adjustment, these customers were able to get what they needed from the city in just a few hours. Previously they had waited days and weeks.
In designing the new processes around the design principles, the first thing needed was to make sure all the required information was there. There were 14 different permits. One of the first things to work out was who reviewed the applications to see what was needed. The city hired and trained from within five generalist engineers to do the reviews. Each engineer had his own team, and each team was self-sufficient.
The most important Design Principles are:
- Initially design work flow around value adding activities, not functions or departments
- Work is performed where it makes the most sense
- Provide a single point of contact for customers and suppliers whenever possible
- Consider every handoff as an opportunity for error. Have as few people as possible involved in the performance of a process
- If things coming into the process naturally cluster, create a separate process for each cluster
- Redesign the process first, then automate it
- Bring downstream information needs upstream
- Capture information once at the source and share it widely
- Ensure 100% quality at the beginning of the process
- Ensure a continuous flow of the “main sequence” (those activities that directly add value to the customer- nothing should slow the value-added steps)
- Look for places to use or create a “generalist” instead of multiple specialists
- Push decision-making down to the lowest levels that make sense
- Use simulation, practice, or role play to test new process designs risk free
- If your process deals with complexity, then consider using teams. Co-locate the teams. If you can’t do this, then network them
- The people who work in the process should be very involved in the analysis, design , and implementation of improvements
- Create a process consultant for cross-functional processes
It is important to look closely at every handoff, from one person or department to another because each one is an opportunity for errors to come in. Minimize the handoffs. And always remember, don’t automate the as-is process. Improve the processes first, then automate them. Insure 100% quality at the front end because it all starts there.
Madison said that San Jose had asked the developers what they considered a good turn-around time was, and they answered that if they got their approvals within two weeks instead of a month, they would be happy. Using the new generalist teams, San Jose got the permit approval time down to two or three days. This result made the developers ecstatic.
This was first done back in 1995 just as the building boom was ramping up. Just before 2000 Madison checked back with San Jose and found that they were handling double the amount of work they did a few years earlier, and with no increase in staff.
