Usage of Machine Learning to Predict Market Attractiveness in the Context of Internationalization

Purpose: The purpose of this study is to explore the perceptions, views, and opinions of chief technology officers (CTO) of software development firms (SDF) about how and why machine learning (ML) methodologies might be used to support foreign market evaluation and selection decisions.Design/methodology/approach: A qualitative research was conducted. The research design is a multiple case study with six semi-structured, in depth interviews with CTOs of SDFs and corporate documents about ML applications from the case study firms as sources of evidence.Findings: The results of this multiple case study suggest the following four findings: 1) The usage of ML to support foreign market evaluation and selection decisions has the potential to improve quality and efficiency, 2) data availability is a key factor of ML to support foreign market evaluation and selection decisions, 3) “easy to use” and “easy to interpret” machine learning supervised methods are the most suitable to support foreign market evaluation and selection decisions, and 4) existing ML development methodologies can be applied to support market evaluation and selection decisions. These findings have a limited generalizability due to the research methodology and are valid only for these case study firms.Research/practical implications: The results of this study might be relevant for researchers who are interested in a further digitalization of decision-making processes. The results might also be relevant for practitioners to better understand the use of ML methodologies in complex and financially important decision-making processes like the evaluation and selection of foreign markets.Originality/value: This work integrated fundamental theories of internationalization based on the works of Johanson and Vahlne in the Uppsala Internationalization Process Model with the concepts and methodologies of machine learning, whose relationship is yet not covered by the academic discourse