The generation of syntactically and semantically valid tests in automatic GUI testing is a key challenge. The inability to generate proper inputs may compromise the capability to reach some areas of the GUI and thereby exercise some functionalities of the application under test. Moreover, easily interpretable tests may facilitate debugging and test maintenance. This talk will discuss this challenge and present how it can be addressed by reusing artifacts and knowledge that is readily available. In particular, the talk will present two main reuse scenarios: migrating meaningful tests across apps and reusing meaningful data present in data sources designed for alternative usage during the testing process.