Agile transformation is a long journey for large companies. Holding off on getting started until all the necessary enablers are in place for successful adoption means the valuable learning which comes through experimentation will be lost. During this formative time, teams will have to cope with constraints which hamper how far down the agile delivery continuum they can operate.
An inability to dedicate primary roles on teams is normal and it is reasonable to start an agile journey with this impediment. However, if nothing is done to address the underlying root causes such as a continued belief in the productivity benefits of multitasking or a lack of understanding of how much work can be done concurrently based on resource capacity then delays, the waste of context switching, and higher defect volume will persist.
Environment or technology constraints might prevent teams from completing all stages of delivery for work items. Phase-based life cycles supported the model of shared testing environments which could be booked by teams for specific periods of time. A shift to end-to-end testing throughout the life cycle will be hampered by a lack of dedicated environments. This forces teams to work in a “Scrum-fall” manner which prolongs launches and will increase the cost and schedule risks of delayed defect detection and resolution. The tactical fix might be to throw money at the problem by provisioning sufficient additional virtual or physical environments, but a more lasting solution might require a shift to a partial or full product/capability/value-stream focus from the current project-centric one.
A lack of high coverage automated testing is a common blocker for teams working with legacy applications. Without this, the cost of testing through the life cycle increases dramatically as does the likelihood of missing regression defects. Investments in developing full automation for an existing application are extremely costly and are rarely justified unless there is a significant backlog of enhancements to be delivered over long product lifetimes. But unless there is a real commitment to empower teams to automate test cases from the very first release for new applications, this situation will never improve.
Constraints and compromises are common when undertaking an agile transformation. But not addressing the underlying root causes will significantly impede the ability to achieve sustainable benefits.