As shown in FIG. 7C, the process begins with the MPS 100 capturing a voice input (block 772) via a network microphone device, such as one or more of the NMDs 103 shown in FIGS. 1A and 1B. The MPS 100 may then transmit one or more messages 782 containing all or a portion of the captured input to one or more remote computing devices associated with a VAS, such as remote computing devices 106a associated with VAS 160. The transmitted voice input may include the wake-word portion (or a portion thereof) and/or the voice utterance portion (or a portion thereof). As discussed above, in some embodiments the MPS 100 selects an appropriate VAS from a plurality of VAS options based on commands and associated command criteria in the set of command information 590 (FIG. 5A). For example, in some embodiments, the MPS 100 selects the ALEXA VAS when the voice input is, e.g., “Alexa, play some INXS,” or selects the GOOGLE VAS when the voice input includes the same voice utterance but a different preceding wake word, such as “Hey Google, play some INXS.”
