The dashboard looked healthy. Average cycle time was down to 3.8 days, lead time had dropped from 14 days to 11, and the team presented both in the quarterly review with some pride. Two weeks later the product manager asked a simpler question: “Is 12 June still safe?” Nobody could answer it from either chart.
Lead time vs cycle time: what each one measures
Lead time is the elapsed time from the moment a request is made to the moment it is delivered. Cycle time is the elapsed time from the moment work starts to the moment it is done. Lead time includes the wait in the backlog, so it is the customer’s view. Cycle time starts later, so it is the team’s view of how long active work takes. Both are useful flow measurements, and both are calculated from issues that have already finished.
That is also their limit. They describe what happened to completed work. A committed release date is a statement about work that has not finished yet.
A vote is a snapshot, and so is an average
The release date was set by a confidence vote, and that vote is expert engineering judgment. It is the best input a team has at planning. But it is a snapshot, and it begins aging as soon as it is taken. The three things that erode a committed release date are scope creep, dependencies surfaced too late, and capacity changes. None of them shows up in an average of past cycle times.
- Scope creep: forty more issues in the Fix Version do not change how long one issue takes. Cycle time stays at 3.8 days while the remaining work grows.
- Late dependencies: an issue blocked by another team’s API has a long wait but no cycle time yet, because it has not started. It is invisible until it finally finishes.
- Capacity changes: an engineer moved to half-time lengthens the calendar, not the cycle time of the issues that do get done.
An average is also a single number, and a release date is a distribution. Cycle times are skewed: most issues finish quickly and a few take much longer. Multiplying the mean by the number of remaining issues gives a date that is right about half the time at best. For why a count of finished work needs a throughput window rather than a mean, see which throughput your release forecast should trust.
Use the flow metrics as evidence, then forecast
Lead time and cycle time are good evidence about whether the system is healthy, and the team should keep tracking them. The step that is missing is the one that turns that evidence into a committed date with a confidence level, and keeps it current.
- Write down the committed date and the confidence it was committed at (P50, P85 or P95).
- Point the release at one Fix Version, so scope has a boundary.
- Choose a throughput window that includes the team’s normal mix of support work and interruptions.
- Record known time off per person and import public holidays per team.
- Re-run the forecast whenever the Fix Version, the issue links or the capacity change, and read the gap in days.
- When the gap turns amber, use the cycle-time chart to ask why. Is work waiting, or is there simply more of it?
An illustration, not customer data. A release is committed for 12 June at 85% confidence. Average cycle time is 3.8 days and stays at 3.8 days for the next six weeks. In those weeks, fourteen issues are added to the Fix Version, two issues become blocked by an external dependency, and one engineer drops to half-time. At the vote the 85% forecast was 11 June. Now it is 24 June, twelve days past the commitment. Cycle time reported no change, and the date moved twelve days.
With the number on the table six weeks out, the team can choose: move some Should-haves to the next release, unblock the dependency earlier, or renegotiate the date while others can still re-plan. The decision is the team’s. The forecast just makes sure it is made before the slip, not after.
Where Advanced Release Planning fits
Advanced Release Planning for Jira does not replace lead time or cycle time reporting. It adds the forecast that sits beside them. It reads the Fix Version, the Jira issue links, the throughput window you choose, and the Capacity Awareness settings for time off and public holidays. It re-runs a Monte Carlo simulation as those change and reports the committed date against the 50%, 85% and 95% bands, with the gap in days attributed to scope, dependencies or capacity. Green at or above P85, amber between P50 and P85, red below P50. The team sets the assumptions: which throughput window, whose capacity, what is in scope. The app does the arithmetic at Monte Carlo scale and keeps it current.
For another chart that shows drift without pricing it, see why the cumulative flow diagram shows the drift but does not price the date. For exactly what the app reads, see the lead time, cycle time and the committed release date reference doc.
Key takeaways
- Lead time runs from request to delivery. Cycle time runs from start to done. Both describe finished work.
- A committed release date concerns unfinished work, so it needs a forecast, not an average.
- Scope creep, late dependencies and capacity changes can move the date without changing cycle time at all.
- Keep the flow metrics as evidence, and read the gap in days between the committed date and the forecast alongside them.
FAQ
What is the difference between lead time and cycle time?
Lead time is the time from a request being made to delivery, including waiting. Cycle time is the time from work starting to work being done. Lead time is always equal to or longer than cycle time for the same item.
Can I forecast a release date from average cycle time?
Not reliably. An average hides the spread, and it does not react to added scope, blocked work or reduced capacity. A throughput-based Monte Carlo forecast gives a date with a confidence level instead of a single guess.
How do I know the committed date still holds if cycle time is stable?
Compare the current forecast at the confidence the date was committed at with the date itself. The gap in days, and whether it reads green, amber or red, tells you.
Track the flow. Forecast the date. See the committed date and its live confidence band in Advanced Release Planning, or start a free trial on the Atlassian Marketplace.




Leave a Reply
Your email is safe with us.