Introduction

“Time isn’t the main thing. It’s the only thing.” - Miles Davis (Jazz Musician)

Decisions to make

• Decide what unit will be used to estimate the capacity of the teams in the program and the size of the features (and epics) in the backlogs — either points or team-weeks (e.g. hours)
• Decide on reasonable expectations for the capacity/velocity of the teams in the program for the next planning cycle (usually based on historical trends)
• Decide how the team-level membership and individual and team-level time allocations will be updated and maintained in Jira Align (including the typical hours per day spent by each team on the backlog)
• Decide what historic data will be used to further define stories and support finer-grain tracking by the teams in hours
• Decide whether demand of particular skills must be balanced against the known, constrained capacity of those skills within the program (i.e. a refinement of the FTE demand)

Context

• Sometimes program leadership teams want to estimate features in points, but the portfolio leadership team wants to estimate portfolio epics in member-weeks (choose member-weeks in these cases)
• Sometimes the teams in the program have been executing for a while and have stable velocity trends, but sometimes teams are new to agile and velocity varies widely from sprint to sprint
• Sometimes there is a mix of scrum teams and kanban teams in the program, and in these cases we still want to understand available capacity
• Sometimes there are specialized skills in the program that can constrain the ability to build plans
• Sometimes the expected amount of paid time off or holiday time should be factored into capacity planning
• Sometimes a significant percentage of the teams' capacity will need to be reserved for expected incoming defect repair (or other emergent work that is not known up front during planning)

Why we care

• Understanding capacity improves planning, better planning improves predictability, and better predictability improves trust and makes work-life balance more sustainable
• Program leaders should strive to bring the “right” amount of feature-level work to planning events based on their understanding of the feature sizes and the program's capacity

The Atlassian view

• Jira Align configures the unit of estimation (e.g. points or member-weeks) at the portfolio level, meaning all programs in the same portfolio will share the same configuration for estimation
• The configured unit of estimation (e.g. points or member-weeks) is applied to both features and portfolio epics (i.e. estimates on both work item types use the same units), while stories are always estimated in story points
• Program leaders collect estimates on features in order to better forecast work and run more effective planning events