TAE promotes selecting automation interfaces that maximize speed, robustness, and functional coverage while minimizing unnecessary UI traversal. For a thin client architecture, most business logic and system functionality resides on the server. To cover functionality efficiently, tests should interact as close as possible to where the logic is implemented—typically via server-side interfaces (e.g., APIs/services, backend endpoints, message interfaces). This reduces GUI overhead and accelerates execution while improving reliability. For a fat client, substantial logic resides on the client side; server-side automation alone may miss critical client behavior, validations, local processing, and UI-driven flows that embody much of the functionality. In such cases, client-side automation (often UI automation or client-level interfaces) is more directly aligned to achieving high functional coverage. TAE also highlights that the “best” interface depends on where behavior is implemented and which interface yields the most stable, fastest checks for the targeted risks. Therefore, the optimal combination is server-side automation for SUT1 (thin client) and client-side automation for SUT2 (fat client), which best meets the goal of broad coverage with minimal execution time.