Kanban system provides organizations simple but powerful metrics that can be directly correlated to business benefits. Metrics in Kanban focus on measuring “time to value” and hence using these measures for continuous improvement generates direct business value.
1) Cumulative Flow Diagram
Cumulative Flow Diagram is a simple but powerful metric that provides rich information at a glance.
The various color bands indicate the work in progress in each of the steps of the value stream. The top most band is the “Backlog” where work items originate and the bottom layer is the archive where they are stored when completed. The following are the intuitions from the CFD
- Lead time and cycle time that are measured along the horizontal axis indicate the expected time for a work item to be completed. Increasing thickness of bands indicate a trend of increasing lead time and cycle times. Reducing the thickness of the bands – by reducing WIP (Work in Progress) or reducing wait time / blocked time will directly affect the time to value.
- Increasing thickness of any individual band indicates a bottle neck to the flow that is being created by the very next step in the process. This triggers a Kaizen discussion amongst the team to remove the bottleneck.
- WIP which is measured along the vertical axis directly impacts the time to value – Cycle time and Lead time. Reducing WIP in conjunction with balancing WIP limits for a smooth flow provides for the most effective business value realization.
2) Control charts
What is a control chart?
Every process varies. If you sign your name five times, they will look similar, but no two signatures will be exactly alike. There is an inherent variation, but it varies between predictable limits. When you are signing your name if you get bumped, you get an unusual variation due to what is called a “special cause”.
There’s also “common cause” variation. Consider a Tennis player. If he has good control, most of his serves are going to be where he wants them. There will be some variation, but not too much. If he does not have control, his serves are not going where he wants them; there’s more variation. Here there is no special causes – no wind, no change of ball – just more “common cause” variation. This results in loss of the serve and easy points for the opposition and may cost him the game – expensive! Likewise, in most processes, reducing common cause variation saves money.
Control charts help visualize this variation. Control charts have the following information
- A centerline, usually the mathematical average of all the samples plotted.
- Upper and lower statistical control limits that define the constraints of common cause variations. These are normally drawn at a distance of 3 sigma (Normal distribution) from mean
- Performance data plotted over time.
How do you use a control chart?
When a process is in control, since data is normally distributed, 99,7% of the data will fall within the 3-sigma control limits. When data points fall outside of this control limit, analysis is done to identify and eliminate data due to “special causes”. Then by further process improvement even common cause variation is reduced which often leads to substantial business benefits.