It's a long read, but I'm hoping others are trying to leverage their investment in Portfolio, and are willing to engage in this topic.
Modeling plans and resources in Portfolio
Background
Active Projects
- We release many Products (>10)
- We run many Projects concurrently
- Active Projects may result in one or more Products being updated
- We coordinate the release of Products amongst the active Projects
- When we release a Project, there is final release activity that is done for the Product.
Planned Projects
- We groom the backlog of Proposed Projects with stakeholders to determine the relative priority
- We assign people to proposed projects to build a staffing plan to let the business know what our Capacity is (vs Demand.)
Staffing
- A person may work on one or more active Project
- We look across our and planned projects to see our staffing utilization.
Current Understanding / Attempts
Projects When many Projects affect many Products, we would like to know when a specific Project is jeopardizing the release of a given Product. “Release” dependencies in Portfolio are limited to a Project/Board – it does not seem possible to make Projects depend on one another.
Option 1 – Create a cross-functional release across for the Product
- This mechanism creates a “release” in each dependent project.
- This model does not work well as a Project may contribute to many Products
Option 2 – Put fixed dates on all projects, if any of them show “Red” - the Product releases they contribute to are in Jeopardy
- Issue – given the number of projects and products we support it’s difficult to understand the implications.
Option 3 – Task level dependency between Projects ( this is a more generalized implementation of https://community.atlassian.com/t5/Portfolio-for-Jira-questions/How-to-make-portfolio-repect-the-release-order/qaq-p/742666
- Stories/Tasks can be defined between Projects to “block,” providing the ability to create dependencies between projects
- Unless its possible to force a task to be the last task in a project, one must define a separate release and sequence it at the end of the Project within the Project releases
- Such tasks are always scheduled for a full sprint, forcing a 1 sprint gap in the dependency chain.
- This is complicated.
Other Options?
Staff Loading
Staff loading across projects seems to be a missing feature – we’d like know who is over/under-utilized based the hours we enter per project. There appears to be no way to set the global availability of a person, and then determine if the assumed allocation across projects is within the global availability.
Options 1 - create a custom report from the underlying database – what are the options? How do we get insight into the schema?
Other Options?
