liveBook · Manning

Cycle time

As explained in Build High Value Software Systems

The elapsed time required for work to move through a production or delivery process.

Cycle time is the elapsed time from starting work on a change to completing it. In a software workflow, the clock might start when a ticket moves to “In Progress” and stop at deployment. Agree on those boundaries before comparing changes: measuring first commit to merge is not the same as measuring ticket pickup to production. Cycle time includes review, CI, and other delays that occur after work starts; it is not just hands-on coding time. section 9.2.2 section 9.5

Lead time starts earlier, when the request is created. The difference between lead time and cycle time shows how long work waited before anyone started it. Activity time narrows the view further to hands-on progress. Keeping these measures separate helps distinguish a backlog problem from delays inside execution. section 9.2.2 section 5.8

A five-day change in a twenty-day journey

If a request takes 20 days to reach completion but work starts only five days before delivery, lead time is 20 days and cycle time is five. The other 15 days were spent waiting to start. To investigate the five-day cycle, reconstruct its stages: coding, CI, review, reruns, and deployment. A long review wait points to a different intervention than slow coding. section 9.2.2 section 5.8

To use cycle time as a flow diagnostic:

  • Choose representative completed changes. Record request, work-start, and delivery timestamps from tickets, commits, and deployment records. section 9.5
  • Break out the longest delay rather than assuming coding is the constraint. Check review queues, CI, flaky tests, and batch size. section 5.8 section 9.2.2
  • Make one small change, then measure subsequent work. Check whether the delay shrank or merely moved to another stage. section 6.6
Figure 9.5 Cycle time measures the span from when work started to when it is merged/deployed. Everything before "Work Started" is queue time, captured by lead time but not cycle time.

The timeline marks the work-start boundary: earlier backlog waiting belongs to lead time, while the span through merge or deployment is cycle time. section 9.2.2

How Build High Value Software Systems covers it

How deep it goes
Defines start and finish boundaries, distinguishes adjacent measures, and gives a timestamp-based measurement procedure.
Examples
Uses a 20-day lead time with five days of cycle time and an exercise tracing recent completed changes.
What it assumes
Assumes familiarity with tickets, commits, and deployments while defining the timing measures and their boundaries.
What you can do after
Diagnose/debug delivery flow by measuring completed changes, finding the longest delay, and checking the effect of one intervention.
How it connects
Connects cycle time to lead time, queue time, activity ratio, batch size, and control charts.

Also explained in

Compare how 5 books explain cycle time →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Build High Value Software Systems. The section links point to the parts of the book it draws on.