Evaluating the Implementation Cost of Blockchain in Organizations through System Dynamics

One of the main obstacles to the adoption of blockchain is the cost of its adoption. The application of this technology in an organization requires the costs of development, design, maintenance, hardware, software, and energy consumption according to its adoption rate. This study uses system dynamics (SD) and machine learning (ML) methods to predict the final cost of blockchain implementation. Compared to mathematical programming, simulation techniques for estimating costs are scarce. However, SD modeling is suitable to account for the complexity and dynamic of systems and support long-term, strategic decision-making. To better understand the system behavior, it is necessary to formulate the relationships between the variables and simulate the values of the variables over time. The relationship between these variables is analyzed using the qualitative SD modeling method with stakeholders through questionnaires and 15 interviews. After identifying the variables, their effect on each other and the implementation cost are investigated. Since the charts obtained from the SD give us the behavior of state and flow variables for time, linear regression applying cross-validation, as one of the ML methods, is used to get a graph showing the system's state as a rate function. Thus, this research provides a reasonable basis for estimating the cost function of blockchain implementation. The validity of the suggested method's results is investigated through sensitivity analysis. The results demonstrate the effectiveness of the proposed model. Simulation results indicate that implementing scenarios such as changes in the average block creation time significantly enhances transaction cost, hardware cost, and software cost, leading to increased implementation cost of blockchain for organizations. The results of this research can significantly help decision-makers develop and apply blockchain technology in organizations.