The retrospective was a good one. The team named what slowed them down, voted on three improvements, and agreed to try a smaller work-in-progress limit and to stop taking support tickets on Fridays. One developer mentioned that she would rotate to the platform team for a few weeks. The sticky notes were photographed. Everyone left feeling the team had gotten better.
Nobody asked the question the room was quietly answering: what did this sprint, and everything we just agreed, do to the committed release date?
What a sprint retrospective is for
A sprint retrospective is the Scrum event where the team inspects how it worked during the sprint (people, interactions, process and tools) and agrees improvements to try next. It is the team’s own continuous improvement loop, and it should stay that way. It is not a status meeting, and it is not the place to defend a date.
That is exactly why the date drifts past it. The retrospective produces action items, and action items are intentions. None of them says how many days they move a release.
The retrospective is where erosion shows up first
At PI planning or release planning, the team gave a confidence vote on a date. That vote was expert engineering judgment: the people who know the codebase, the dependencies and each other looked at the plan and said “we can commit to this.” It was right on the day it was taken.
Afterwards, the three things that erode a release date start working, and the retrospective is usually the first place the team talks about all three:
- Scope creep. “We keep absorbing small asks mid-sprint.” Everyone agrees it hurts. Nobody counts how many days the asks added to the release.
- Dependencies surfaced too late. “We were blocked on the identity team for four days.” It is logged as a lesson learned. It is also a sequencing change the vote never saw.
- Capacity changes. A developer rotating to another team, a support rotation, a holiday week. Throughput moves, and the date moves with it.
The retrospective records each of these as a discussion point. What it does not do is convert them into days against the committed date. So the release slide stays where it was until the sprint where it cannot.
Add one question to the end of the retrospective
You do not need a new meeting, and you should not turn the retrospective into a reporting session. You need one closing question, answered with evidence:
“Given what we changed and what we learned, how confident are we now in the committed date, and how many days are we from our 85% date?”
Answering by hand is where it falls apart. Re-running a forecast means re-reading throughput, recounting remaining scope, adjusting for who is actually available, and repeating the simulation thousands of times to get a real range. Teams skip it because it takes an afternoon, and the retrospective lasts an hour.
The team owns the assumptions. The forecast does the arithmetic.
This is what Advanced Release Planning for Jira is built for. A confidence vote is a snapshot; Advanced Release Planning keeps it live. The team decides the inputs: which throughput window reflects how it works now, whose capacity counts, what is in scope for the Fix Version. The app runs Monte Carlo simulation over the team’s real Jira throughput and holds the committed date against a confidence band (50/85/95%), so the answer is on screen before the retrospective ends.
In practice the last ten minutes look like this:
- Update what changed. If the team agrees the last six sprints no longer reflect how it works, it picks a new throughput window. If a developer is rotating off, her capacity changes. If scope was added, the Fix Version shows it.
- Read the band. “Committed date is June 7. Confidence at June 7 is 71%, down from 85% at the vote. P85 is June 14.”
- Decide while it is cheap. Cut or defer scope, resequence the dependency, or renegotiate the date with stakeholders before the slip rather than after it.
The judgment is still the team’s. It is now exercised on current evidence every sprint instead of on the evidence from the day of the vote.
Improvement actions show up later, in the throughput
One honest limit: a retrospective action such as a smaller WIP limit does not change the forecast on the day it is agreed. Its effect shows up in throughput over the following sprints, and the forecast picks it up as the team’s completed work changes. That is the right behavior. A forecast that gave credit for intentions would just be a more elaborate way of hoping. Pairing this with the sprint review gives you two recurring points each sprint to read the band: once with stakeholders, once with the team.
Keep the vote live
The retrospective makes the team better at delivering. The committed date still needs someone to check whether delivering at the current pace keeps the promise. Read the band, name what moved it, and renegotiate before the slip. How the forecast is built and what the app does and does not read is in the sprint retrospectives and the committed release date feature doc.
Bring a live confidence level to your next retrospective →




Leave a Reply
Your email is safe with us.