The Functional Resonance Analysis Method (FRAM) has been pointed out by several prior studies as an effective approach for modeling socio-technical systems. Despite previous contributions, FRAM still suffers from limited tractability in case of large models such as a difficulty of carrying out reliable and insightful analysis. This paper addresses this gap by proposing an approach inspired by network theory to enhance traditional FRAM analyses, namely the functional random walker (FRW). The definition of FRW complements the notion of random walks accounting for the variability being transferred inside the system from one function to another, offering a cost-effective means of exploring FRAM instantiations. Five design criteria are presented to reproduce such analysis with any FRAM model, along with two indicators, namely the probability of visiting nodes and the mean passage time. These two metrics permit to study the FRAM model both at function level (how a single function behaves within the model) and at system level (how multiple functions interact with each other). A walk-through application of FRW is presented for a healthcare setting. This application revealed how the FRW sheds light on systems’ strengths and vulnerabilities, and highlighted underlying knowledge, otherwise hidden in cluttered visual representations of FRAM instantiations.