An electronic document having a current layout can be presented in an application user interface. Based on a detected interaction with an object in the electronic document, an intent can be determined. The determined intent can be used to identify different sets of commands that result in corresponding visual representations that are likely to satisfy the determined intent. Different activation regions can be defined within the electronic document. A different command set, each identified from the determined intent, is associated with each different activation region. Movement and release of the selected object to a position within an activation region is detected, and the command set associated with the activation region is executed to perform a multi-step operation that modifies at least the current layout of the electronic document to result in the corresponding visual representation.