RPAEstiMate: effort estimation model for robotic process automation transformation

Abstract

Purpose This study aims to develop an adaptable effort estimation model for Robotic Process Automation (RPA) transformation projects. This model can assist RPA project teams in calculating the work required for the successful completion of RPA transformation initiatives. RPA operates as a virtual bot that uses rule-based business processes. Design/methodology/approach This case-based study was conducted in collaboration with a company facing challenges in identifying an adaptable effort estimation model for RPA transformation. The research progressed through four stages. In Stage 1, the company’s existing estimation method was analysed. Stage 2 focused on identifying key factors influencing effort estimation in RPA projects through a pre-survey and a literature review. In Stage 3, a tailored effort estimation model for RPA transformation was developed. Finally, in Stage 4, the model was validated for model accuracy and effectiveness in predicting effort requirements for RPA transformation. Findings The authors developed the RPAEstiMate model, an effort estimation model, to assess the complexity and resource requirements for RPA transformations using a case-based approach. The model addresses the current gap in the lack of adaptable estimation methods. During the model development, 15 significant factors were identified. These were categorised into Business Process-, Technical- and Development Team-related factors. RPAEstiMate demonstrates 80% accuracy in predicting effort requirements. Research limitations/implications Because the research is conducted within a company in Sri Lanka, the study showcases limitations in scope and generalisability. The RPAEstiMate model satisfies both research and company objectives within this context. The need for more data from various industries limits external validity. Therefore, interviews with SMEs from different sectors were conducted to broaden the insight. Further, future research must use the expanded data set to increase applicability. Practical implications The RPAEstiMate model needed to be customised to work effectively across different geographical regions, industries and different types of business processes. The existing model was created based on data and experiences from only one company. Organisations can benefit from using this model as a starting point while tailoring it to their specific context through iterative testing and validation. Originality/value Based on the online survey analysis conducted during the study, it was identified that out of 53% of respondents who reported conducting effort estimation in RPA transformation, 75% depend on Subject Matter Experts’ (SMEs) prior experience for estimation and only 25% use estimation processes provided by RPA vendors like “UIPath” and “Automation Anywhere”. This study addresses the lack of standard approaches and the reliance on RPA SMEs for effort estimation in RPA transformation projects by developing the RPAEstiMate model, emphasising the study’s originality.