According to the invention, an information output, e.g. a voice output, video information output, or a simultaneous output of video and audio information, which is to be transmitted via a packet-oriented network, is generation by means of pre-coded information output components. Said pre-coded information output components are present in a memory system and are retrieved as part of the processing of a request for the output of information by an information output system. Similarly, generation rules for the information output can be stored in the memory system and accessed by the information output system in order to generate the information output. According to one embodiment, information output components or generation rules are created in an information output configuration system and are subsequently made available in the memory system. The invention optimizes the use of resources for information-output services. The inventive distribution of functions permits the advantageous utilization of the performance of the related components for processing services of this type. The recorded announcement services capability known in TDM (time division multiplexing) networks can thus also be implemented in packet-oriented networks.