Dependencies are a normal part of delivering complex products.
When several teams contribute to the same solution, some work will need to happen in sequence. Teams will occasionally require specialist expertise. Shared platforms, environments, governance functions, and integration points will create legitimate connections between people and work.
The problem is that Agile Release Trains often respond to every dependency in the same way, i.e. by adding another coordination mechanism.
For example:
Let’s say a dependency is placed on a board. RTE schedules a cross-team meeting… aka “sync”. An escalation path is created. The RTE follows up with the teams. Everyone agrees to monitor the situation closely.
These “actions” may help a temporary dependency. But they DO NOT address recurring dependencies.
Before deciding how to manage a dependency, an RTE should ask:
What kind of dependency is this, and what response does that type require?
The answer should lead to one of the following responses: