Strategy Performance Part II 09 December, 2010

Even though most people are concerned about their own health, I'm guessing that a small percentage of those people don't really think about the health of the company they work for. I would argue that although many employees are constantly searching for that next innovation, simple process improvement would improve the average company just as much.

Process improvement is not only about improving, it's also about knowing what is possible. A process is working well if it consistently performs well, and has a low level of daily variance. When looking at a company checkup in terms of its process health, the investment in improvement is not as expensive as first impressions may indicate. In fact, many studies show that it is not the willingness or ability to improve, but the time required that is often the problem.

Four main goals of process improvement are:

• Cutting costs
• Cutting waste
• Improving quality
• Removing stress

Two popular ways of improving processes are benchmarking and waste identification. Either way, it's all about removing things that do not add value and adding more value where it is needed.

Benchmarking is all about using metrics to grade a process. These grades can then be compared to either other internal benchmark grades or external benchmark grades dealing with similar processes. Internal comparisons are used for continuous improvement from where the company was previously at. External comparisons are used to find out how well a process compares to the best versions of those processes on a local, national, or even global scale.

Benchmarking and process improvement is comprised of the following steps:

• Talk to those dealing with the process (to gain information and buy-in)
• Create metrics for comparing improvement levels
• Collect metrics before a change (as the process exists today)
• Make the change
• Collect metrics after a change
• Make sure the improvements and measurements are statistically significant
• Present findings, real world benefits, and recommendations
• Officially incorporate the changes into the process
• When needed, benchmark again for continued improvement

Measurement methods chosen should usually be standard to keep bias out of the question. Some possible measurements include:

• Time
• Number of people involved
• Number of manual steps
• Number of communication methods or problems
• Amount of redundancy

Waste identification attempts to identify for removal all resource expenditures that do not create value. There are eight distinct types of waste that may exist:

• Transportation (moving things more than needed)
• Inventory (any unneeded storage or depreciation)
• Motion (people moving more than needed)
• Waiting
• Over production (making more than needed)
• Over processing (poor tool use)
• Defects (taking extra time to find defects and fix them)
• Underutilized resources

When benchmarking or identifying waste, it is important that precise documentation is kept. Several popular documentation techniques include:

• Process Diagrams & Value Stream Mapping
• Cause and Effect (also known as either Fishbone or Ishikawa diagrams)
• Pareto Charts / Analysis
  

Process diagrams are simply what they say they are. A flow of beginning to end for a process. This includes preparation for the process, the process, and any cleanup. It also includes inputs, outputs, capacities, times, and bottlenecks. It is sometimes a good idea to limit the process to be improved to just a portion of the whole process if it is very complex or big. In this case it is common to Plan, Do, Check, and Act (PDCA). PDCA is really just a business version of the scientific method.

Cause and Effect diagrams are useful for finding root causes to problems. Looking at the following diagram, you can see why some people call it a fishbone diagram. Every time there is a reason for the effect, an attempt is made to find a lower level cause.


Pareto charts are cross plots of specific problem types vs total problems (relative vs. absolute). They are useful in showing which problem areas should be addressed first.
If it is uncertain whether a process needs to be benchmarked or improved, try looking at that process from a customer's perspective and see if you would be impressed or left wanting.