Production Metrics: Cycle Time and Throughput Rate

In addition to Velocity and Predictability, there are other crucial Output metrics, especially relevant in flow-based systems like Kanban: Cycle Time and Throughput.

Cycle Time measures how long it takes for a work item to go through the process (or a part of it), from start to finish.

Throughput Rate measures how many work items are completed in a given period of time (e.g., items per week).

These metrics provide direct insight into workflow efficiency and are intrinsically related to the amount of Work in Progress (WIP).

They are critical to understanding delivery velocity and identifying opportunities for process optimization.

Relationship with WIP (Little's Law)

The connection between Cycle Time, Rate of Throughput and Work in Progress (WIP) is described mathematically by Little's Law.

This law states that, in a long-term stable system, the average WIP is equal to the average Throughput Rate multiplied by the average Cycle Time.

WIP = Throughput x Cycle Time

This relationship has important implications:

• If the Throughput Rate is constant, reducing the WIP will directly reduce the Cycle Time. That is, limiting the amount of parallel work makes each individual item complete faster.
• Alternatively, if you want to reduce the Cycle Time without reducing the WIP, you must increase the Throughput Rate.

Understanding this law helps teams manage their WIP consciously to optimize delivery speed (Cycle Time).

Identification and Elimination of Bottlenecks

Cycle Time and Throughput Rate are key tools to identify and manage bottlenecks in the workflow.

A bottleneck is a process step whose capacity limits the overall system throughput rate.

By measuring the time items spend in each state (Kanban board column), you can identify where work is piling up or where partial Cycle Times are excessively long.

Visualizing this on a Kanban board, especially with WIP limits, makes bottlenecks obvious.

Once a bottleneck is identified (e.g., QA testing, code review, waiting for external dependencies), the team can focus its improvement (kaizen) efforts on eliminating it or expanding its capacity.

Common strategies include:

• Collaboration: Have members from other stages help with the bottleneck (e.g., developers helping with testing).
• Automation: Invest in automated testing to reduce manual burden.
• Process Improvement: Establish clear rules (e.g., response times for code review) or refine the process at the bottleneck stage.
• WIP Man