Critical Success Factors Impacting Intelligent Process Automation— A Data-first Machine Learning Approach

Intelligent Process Automation (IPA) refers to a combination of multiple technologies, including Digital Process Automation (DPA), Robotic Process Automation (RPA), and Artificial Intelligence (AI). IPA produces significant business process automation that improves business value and leads to digital transformation for an organization. In this study, I seek to identify the success factors impacting IPA-led digital transformation.

I argue that multi-level theorizing is necessary to explain the complex patterns of success factors that are critical for IPA success. For this purpose, I propose to follow a multi-step research methodology, consisting of in-depth key informant interviews, decision-tree induction, and theory abduction. A sample of 176 IPA projects from the financial services domain, implemented by a multi-billion-dollar global IT services company, will serve as the data set for my data-first machine learning investigation.

I will draw on several theoretical perspectives and qualitative interviews to identify predictors of IPA success from multiple levels, such as the domain, process complexity, technology complexity, and technology governance. For my first study, I utilize decision-tree induction to examine three dependent variables corresponding to different dimensions of IPA success — Full Time Equivalent Reduction, Process Efficiency Improvement, and Process Accuracy Improvement.
In the second study, I propose to combine these three dependent variables into a formative construct, IPA Success Index, that wholistically captures the extent of IPA success. Decision-tree induction will again serve as the methodology to identify predictors of the IPA Success Index. I will then abduct away from the findings of the first two studies to develop generalizable theoretical propositions.

In the third study, I propose to investigate in-depth the relationships between the key success factor(s) identified in earlier studies and IPA success using econometric analysis. The identified success factors, the IPA Success Index, and theoretical propositions from my dissertation will contribute towards the growing literature on intelligent information systems within the larger stream of IT business value research. The practical benefits of these studies will accrue to managers and organizations across the globe and enable them to maximize the benefits of IPA-led digital transformation.