Short answer: To plan capacity across several agile teams, calculate each team’s available person-days first, count every shared specialist in exactly one place at a time, and only then roll the team totals up into a program view. Skipping the “count once” step is how large agencies commit to more work than their people can deliver.
Flat budgets and workforce reductions have put a lot of programs in the same spot: four teams, one mission, and the same two or three specialists that every team needs. “Fewer people, same mission” is not a slogan here. It is the arithmetic you have to do every sprint.
Why does multi-team capacity planning break down in large agencies?
Each team can plan its own sprint well and the program still fails, because the problems sit between teams. A database engineer, a security reviewer, or a business analyst is listed on two or three rosters at once. Each scrum master sees that person as available and plans around them. Nobody adds the commitments up, so the same person is promised at 200 percent.
Government structure makes this worse. Staff are detailed in from other offices, contractors rotate on and off with contract modifications, and leave and training calendars are managed outside the team. The program office usually sees the problem only after a sprint review shows missed commitments on several teams at once.
How do you plan capacity across several teams?
Use this five-step sequence in the same order every sprint:
- Start at the team level. List every person on each team with the number of working days in the sprint. Remove agency holidays first, so a mid-sprint federal holiday does not surprise anyone.
- Assign shared people once. For each specialist who supports more than one team, split their days across teams so the total never exceeds their available days. Write the split down. If a team lead says “we get her half-time,” that half-time must appear as the other half on a different team’s roster.
- Subtract days off before you commit. Leave, training, required agency events, and on-call duty come out of each person’s days. Do this per person, not as a team-wide percentage.
- Convert to points using each team’s own history. Divide the team’s recent completed points by its recent available person-days to get a points-per-person-day rate. Do not borrow another team’s rate.
- Roll up and check dependencies. Add the team commitments into a program total, then look at which stories need a shared specialist. If two teams both need the same person in the same week, resolve it before the sprint starts, not during it.
What does this look like with real numbers?
Take a program with two teams sharing one database engineer during a two-week sprint. A mid-sprint federal holiday leaves 9 working days.
Team Alpha has five dedicated developers and the shared engineer. Leave and training total 7 person-days. Alpha’s last three sprints averaged 0.95 points per available person-day.
- Counting the engineer once (50 percent to Alpha): 5.5 people × 9 days = 49.5 person-days. Minus 7 days off = 42.5. At 0.95 points per day, that is about 40 points.
- Counting the engineer at 100 percent on both teams: 6 people × 9 days = 54 person-days. Minus 7 = 47. At 0.95, that is about 45 points.
The difference is about 4 points for Alpha. If the second team makes the same double-count, the program has promised roughly 9 points of work that does not exist. Nothing in either team’s plan looks wrong on its own. Only the program-level view exposes it.
Why does a single tool inside Jira make this repeatable?
Spreadsheets can do this math, but they drift. Rosters change, a second copy appears, and the sheet no longer matches the backlog. When agencies move work onto Atlassian Cloud, including Atlassian Government Cloud (AGC), the sensible next step is to keep capacity next to the backlog it constrains.
Sprint planning with Capacity Planning for Jira puts sprint planning and team capacity on one screen inside the Jira backlog. It uses past velocity, days-off, and allocation, so you see what a team can realistically take before you commit. Shared people are allocated in one place, which makes over-allocation visible rather than assumed. The app is Cloud Fortified and runs on Atlassian’s SOC 2 Type II / ISO 27001 platform. It also leaves a record of who was allocated, which days off were subtracted, and why a team committed to N points. That is a planning record, useful when leadership asks how a number was reached. It is not a compliance product.
FAQ
How do you handle a specialist shared across three teams?
Split their available days into fixed blocks per team before the sprint, and record the split. Re-split at sprint planning if priorities change, never mid-sprint without telling the other teams.
Should every team use the same velocity rate?
No. Points are relative to each team’s own estimates. Compare capacity in person-days across teams, and compare points only within a team.
How early should program-level capacity be checked?
Before each team’s sprint planning meeting, so conflicts over shared people are settled with the full picture instead of discovered afterward.
A 90-day rollout idea: in month one, list every shared person and their splits; in month two, add per-person days off to each team’s plan; in month three, review the program roll-up against actual delivery and adjust the points-per-day rates.
Try it: Install free on the Atlassian Marketplace · Help docs




Leave a Reply
Your email is safe with us.