Below are the Throughput (total items made live per week) and Cycle Time (the mean number of days taken an item takes to get done) stats since we started Kanban on one of our teams.
Our stats show several spikes, due to the team being subject various release freezes, with only critical patches allowed into Live.This becomes more obvious when reviewing our State CFD (last chart). To counter this we have started recording data for the Engineering stages only which will give us more useful data to reflect on as a team.
The items flowing through the pipeline are classified as Small, Medium and Large, the stats below are an amalgamation of these, we intend to start producing Cycle Time Metrics for each.
We still record data for Total Throughput and Total Cycle time which in the end is all the customer is interested in (when will I get it!) but having these additional metrics are proving useful for improvement.