Short answer: To manage allocation across government programs, record each person’s percentage split per program before sprint planning, convert it into available hours net of days off, and refuse any commitment that pushes a person above 100%. Over-allocation is not a staffing problem you discover at the end of a sprint; it is a planning error you can catch before you commit.
Picture a state benefits agency after a hiring freeze. The same four engineers now support the eligibility system modernization, a federally mandated reporting change, and the legacy case-management platform that still pages someone at 2 a.m. Every program manager believes they have “the team.” Each one plans a full sprint. By week two, the reporting deadline slips, and nobody can point to the moment the plan stopped being realistic.
Why is over-allocation so common in government programs?
Public-sector staffing rarely maps one person to one program. Staff are funded from multiple lines of accounting, detailed to task forces, pulled into audit responses, or shared between a modernization effort and the operations work that keeps the current system alive. When budgets flatten or headcount drops, the mission does not shrink. Fewer people, same mission means the same names appear on more plans.
The core failure is that allocation lives in one place (a staffing spreadsheet, a resource request, a labor-charging system) while commitment lives in another (the Jira backlog). Nobody adds the two together until the sprint is already underway. Three patterns show up again and again:
- Phantom capacity: each program counts a shared engineer at full time because nobody told them otherwise.
- Invisible operations load: help-desk rotations, patching windows, and incident duty consume time that never appears as a story.
- Untracked absences: leave, training, and federal holidays are subtracted in someone’s head, if at all.
How do you manage allocation across multiple programs?
The following five steps work whether you run two programs or twelve. The goal is a number per person, per sprint, that every program manager agrees on before planning starts.
- Publish an allocation table each sprint. For every person, list the percentage of time assigned to each program and to operations. The row must sum to 100% or less. If it sums to 130%, you have found your problem before it costs you a sprint.
- Convert percentages into hours, not good intentions. Multiply working days in the sprint, minus days off, by realistic focus hours per day (many teams use 5 to 6), then apply each program’s percentage. This is the capacity that program may plan against.
- Reserve operations time explicitly. Treat on-call, patching, and audit support as an allocation line, not as interruptions. If it happens every sprint, it belongs in the plan.
- Plan each program against its share, using its own velocity. Compare the hours or points each program is about to commit with that program’s slice of capacity and its recent delivered velocity. Commit to the lower of the two.
- Escalate conflicts with numbers, not complaints. When two programs both need the same person above their agreed share, bring the allocation table to the portfolio or PMO lead. A specific trade-off (“Program B gets 10 more hours if the reporting change moves one sprint”) gets a decision faster than “we are stretched.”
A worked example: one engineer, three programs
Take one senior developer on a hypothetical two-week (10 working day) sprint. She has one day of annual leave and one federal holiday falls in the window, leaving 8 working days. At 6 focus hours per day, she has 48 available hours.
Her agreed split is 50% eligibility modernization, 30% reporting change, and 20% legacy operations. That gives 24 hours, 14.4 hours, and 9.6 hours respectively.
At planning, the modernization program loads 30 hours of her work, the reporting team loads 20 hours, and operations history shows about 10 hours of incident and patching work per sprint. Total demand: 60 hours against 48 available. She is at 125% before the sprint begins.
With the table in front of them, the leads have options: move 6 hours of modernization work to the next sprint, reassign a reporting task to a colleague who sits at 80%, or formally accept that the reporting change slips. Any of those is a decision. Discovering the 12-hour gap in week two is not.
Allocation decisions are also a record
Agencies are under steady pressure to show what happened, who acted, and why. A saved allocation table with days off subtracted and the committed hours beside it is a plain record of why the team committed to what it did. That is good governance practice on its own; it is not a compliance control, but it makes post-sprint reviews and leadership briefings far easier to support with evidence.
Making it repeatable inside Jira
The steps above fail when they depend on someone rebuilding a spreadsheet every two weeks. As agencies consolidate work onto Atlassian Cloud, including Atlassian Government Cloud (AGC) for those that require it, keeping allocation in the same place as the backlog removes the reconciliation step entirely.
Sprint planning with Capacity Planning for Jira puts this on one screen inside the Jira backlog: each person’s allocation, their days off, and the team’s past velocity sit next to the issues you are about to commit, so over-allocation shows up while you are planning instead of after. The app is Cloud Fortified and runs on Atlassian’s SOC 2 Type II / ISO 27001 platform.
FAQ
What allocation percentage counts as over-allocated?
Anything above 100% of a person’s available hours after days off are removed. Many teams also treat a sustained 90% or more as a warning, because unplanned work always arrives.
Should operations and support work be allocated or left as buffer?
Allocate it. If operations consumes roughly the same time every sprint, an explicit line makes the plan honest and stops each program assuming it gets that time.
Who should own the allocation table?
Usually the PMO or portfolio lead, with each program manager reviewing their column before sprint planning. One owner prevents competing versions.
Try it: a one-sprint allocation check
Before your next planning session, list every shared person, their percentage split, and their days off for the sprint. Add up what each program intends to commit. If any row exceeds 100%, resolve it before anyone commits. Then run the same check in Jira so it carries forward every sprint.
Try it: Install free on the Atlassian Marketplace · Help docs




Leave a Reply
Your email is safe with us.