Software development teams have been struggling hard in identifying the right software model to achieve an effective and efficient system. In this digital era, with immaculate products being expected by the customers, it is imperative to etch a versatile software model for Internet of Things-Fog systems (IoT-Fog) which is expected to be conducive in this century. This paper discusses the results of identifying a reliable software model for IoT-Fog based system by employing Discriminant Analysis. The metrics and models were identified through intensive study and discussion held with research experts that covers both software and IoT-Fog based systems. For this reason, it is extremely significant to look out for an appropriate software model considering the risk factors. The agile process model for IoT-Fog computing is proposed to enhance the reliability when compared with other models using specific metrics. The analysis shows that agile software model will be an optimum solution for IoT-Fog based systems.