Seeking Augmentation of Knowledge Work through Machine Learning: A Case Study in the Field of Financial Planning

Möllers, Miriam; Berger, Benedikt

Abstract

Organizations increasingly use machine learning (ML) techniques to augment knowledge work with data-driven predictions. An example of such knowledge work is forecasting, a resource-intensive and discursive process, in which knowledge workers collaboratively analyze and interpret the business and market data to discuss strategic goals. Through a case study in a financial planning and analysis (FP&A) department of a telecommunications company, we investigate how finance professionals integrate ML forecasts within existing, manual forecasting activities. Using the lens of epistemic cultures, we interpret the manual forecasts as key objects through which finance professionals create a shared understanding and form assumptions about the current and future business influences. The ML forecasts introduce new insights to these discussions, offering different interpretations and meaning. Consequently, the finance professionals struggle to make sense of the differences between the ML and manual forecasts, to evaluate their quality, and thus, to fully integrate them in their activities.