Measure cycle time without calling every unrecorded minute waiting
A task can take two hours while recording half an hour of active effort. The difference alone does not show that anyone waited for ninety minutes. Define events and states before using the number to make decisions.
The destination is in Spanish. A waitlist does not guarantee immediate access or a free plan.
Situations to review
Fictional item A starts at 09:00 and finishes at 10:00, with 20 minutes of recorded, non-overlapping active effort.
Measure 60 elapsed minutes between the declared events and report the 20 active minutes separately. Look for recorded states to explain the remaining intervals.
Without evidence of their states, the remaining 40 minutes stay unclassified. Subtraction proves neither waiting nor its cause.
Fictional item B starts at 09:00 and finishes at 11:00, with 30 minutes of recorded, non-overlapping active effort.
Report 120 elapsed minutes and 30 active minutes using the same calendar as item A. Do not substitute duration for effort or attribute the remainder to a person.
Without explicit states, the other 90 minutes remain unclassified. They do not establish savings, spare capacity or low productivity.
Fictional item C starts at 09:30 and remains open at the 11:00 cutoff.
Record an observed age of 90 minutes, its open status and the cutoff. Without activity records, do not assign effort or a distribution of states.
Do not present age as a finished cycle or silently mix it with completed items. Report open work separately so accumulating time stays visible.
How to evaluate the work
- Declare the events defining each measure and the unit of work. Lead time usually runs from request receipt or logging to completion; cycle time runs from work starting to completion. Configuration matters: Jira’s Control Chart uses selected workflow columns and can include further work after reopening. State the method rather than assuming one universal definition. A selected status is not proof of completion: verify the completion event or filter before calculating completed-item statistics.
- Define the date, timezone, elapsed-clock or working calendar, pauses and reopening rules. For this method, active intervals must be within the observation window and non-overlapping. Waiting needs an explicit recorded state: unclassified gaps are not automatically waiting, and duration alone does not explain their cause. Parallel person-hours use a different effort unit and cannot simply be subtracted from elapsed-clock minutes.
- Separate completed and open items. Show the count, mean and median with the method; the center of a small sample is not a universal target. A Think² collaborator can organize sources and proposals with the computer they can have, groups and context, a compatible model and authorized tools; no specific measurement integration is established. External sharing requires approval. Claims of improvement need comparable periods and populations and a prior baseline.
- Template: item; events/date/timezone; start; finish or cutoff; open/completed; calendar; elapsed time; active intervals; remaining states; statistical inclusion. Fictional example on the same date and timezone, elapsed-clock minutes: A 09:00–10:00 completed, 60 elapsed, 20 active, 40 unclassified; B 09:00–11:00 completed, 120 elapsed, 30 active, 90 unclassified; C 09:30–11:00 cutoff, open, age 90, effort unknown. For completed A and B only, cycle-time mean=(60+120)/2=90 minutes and median=90; active effort totals 50 minutes, with mean and median of 25. C is reported separately, not included as a finished cycle. Without a creation or request timestamp, lead time cannot be calculated. This arithmetic was checked locally on fictional values; it does not establish customer results, integration, savings or capacity.
Frequently asked questions
Are cycle time and lead time the same?
Usually not. Declare the exact events: request or logging to completion differs from work start to completion. Workflow states and reopening rules can affect the calculation.
Is elapsed time minus active effort waiting time?
Not from subtraction alone. You need events identifying that state. Otherwise report unclassified time. Even a recorded state does not, by itself, establish the cause.
What about work that is still open?
Report its age at the cutoff, count and status separately. Do not treat age as a final duration. Explicitly excluding open work from completed-item statistics does not imply the whole cohort will finish with that mean.
Sources and scope
- Atlassian — View and understand the control chart — Definitions in Jira, selected workflow states and reopened work; not proof of integration or a universal calendar.
- GOV.UK — How to set performance metrics for your service — Baseline, monitoring and segmentation; not a universal cycle-time median rule.
Keep exploring
The destination is in Spanish. A waitlist does not guarantee immediate access or a free plan.