Alternate and exception flows, preconditions and postconditions, keeping use cases at a consistent level of detail, relating them to test cases, and knowing when a use case is the wrong tool.

For readers whose use cases are now the basis for building and testing, where a missing path becomes a defect.

You come away able to document the paths users actually take rather than only the intended one, write use cases a tester can work directly from, and keep detail consistent across a set of them.

Specifying behaviour across system boundaries, failure and recovery paths, non-functional requirements stated so they can be verified, and modelling behaviour that varies by state.

Written for people specifying systems that integrate with others, where behaviour depends on something outside your control.

You can specify what happens when an external system fails, write a non-functional requirement precise enough to test, and describe behaviour that changes depending on state.

Choosing a specification approach to fit the work, where use cases become unwieldy, alternatives and their trade-offs, and setting standards that hold across several teams.

The deepest treatment, questioning when this style of specification helps and when it produces documents that are never read.

You can choose a specification approach suited to the work rather than to habit, recognise when use cases have become a liability, and set a standard that holds across several teams.