A New Copy!

Today we tackle Chapter 2 of  Daniel S. Vacanti’s Actionable Agile Metrics for Predictability. Chapter 2 is titled The Basic Metrics of Flow.  The concept of flow is critical to predictability.   Buy your copy today and read along!

The basic metrics of flow (noted in Chapter 1) are:

  1. Work in progress
  2. Cycle time
  3. Throughput

Work in progress (WIP) is based on the central premise that work arrives, gets completed and then departs.  Value is recognized in the equation when the work leaves the process.  WIP is useful because it is an excellent way to monitor and predict system performance. For example, as more work is started and not completed (WIP increases), system performance falls because less work (and value) will be delivered. A second reason WIP is important is that the other flow metrics are built using WIP.

WIP is defined as the amount work that has arrived to be worked on in a system and has not yet exited the system, regardless of whether the item is actively being worked on or being delayed. The hard part is deciding on a definition of when work arrives.  Vacanti points out that the definition of arrival is easier in a pull system than it is in a push system.  In a pull system, such as Kanban, work arrives when the team pulls it from the backlog.  In push systems, where work is assigned to the team, defining ‘arrived’ is typically is harder to define. My experience is that teams in a push environment typically have higher WIP than in pull environments because saying yes and starting work is easier than telling people that work will need to wait until later. In general, there is less consideration of team capacity in push methods than pull, because pull methods only take work when they have the capacity to move it forward.

Vacanti defines WIP as the number of “discrete units of customer value that have entered the given process, but have not exited.” There is no acknowledgment of size or complexity. Size and complexity have no impact on the predictive nature of the metric. (Chapter 3 is on Little’s Law which we discussed in September 2014)

Cycle time answers the question asked in Chapter 1: When will “it” be done? Cycle time is defined as the amount of elapsed time that a work item spends as work in process. Cycle time is a direct reflection of the calendar, which is the one element every customer understands. Cycle time (also called lead time or flow time even though there might be slight differences in the definition) includes ALL of the calendar time between starting and completing.  All delays, regardless of the reason, is a waste. Delay is the enemy of flow.

Cycle time can be used as a fairly good proxy for predicting cost.  The longer a piece of work takes to complete, the more costly it will be because in software labor is often the largest cost.  

A secondary metric mentioned by Vacanti is flow efficiency.  Flow efficiency is the ratio of total elapsed time that an item is actively worked on compared to the to the total time it takes for something to be completed. Any reduction in the amount delay (delay is the time an item is inactive/waiting/on-hold) will by definition improve overall flow efficiency.

Throughput is a measure of the number of items that transverse the process in any given period. Throughput can be thought of as departure rate; that is how many work items complete and leave the process in a given period. Vacanti notes that arrival rate (departure rate’s mirror image) can also be an important metric.  If arrival and departure rates are not synchronized, WIP and cycle time will increase.

The Chapter concludes with a list of key learning and takeaways which acts as a valuable summary.  The last takeaway notes that the three metrics are the basis Little’s law which foreshadows Chapter 3.  

Previous Installments:

Week 1: Introduction and Game Plan
Week 2:  Flow, Flow Metrics, and Predictability

Actionable Agile Metrics for Predictability: An Introduction by Daniel S. Vacanti


Get your copy and begin reading (or re-reading)!