In an era of globalization, cross-fertilization of technologies and industries, and changing markets, firms are introducing new ways of creating or capturing value through Business Model Innovations (BMI). In recent years, BMI has become one of the priorities of practitioners, and has attracted the interest of scholars since product or process innovations on their own are perceived insufficient in the current internet era when other sources of competitive advantage are being needed. However, BMI can be difficult to manage for many firms, and despite increasing debate in the field, there is a lack of understanding about how BMI processes unfold. The purpose of this thesis is to explore BMI processes in multiple industrial and organizational contexts. To achieve this, the thesis is based on four papers written during the course of this PhD research which draw on empirical studies of diverse industries such as manufacturing, automotive, construction, publishing, and home furnishing. The firms studied in this thesis are new ventures developing new Business Models (BMs), Small and Medium-sized Enterprises (SMEs), and multinational corporations that have been working with BMI, either in parallel or as a substitute to their existing BMs. The empirical observations support the distinction of two approaches to BMI: purposeful and unintentional. Purposeful BMI tends to be planned and starts with attentive cognitive search for a new BM, including recursive conceptualization, creation and offline evaluation of alternative BMs. The process is followed by experiential learning and adaptation of the new BM. Unintentional BMI refers to the emergence of a new BM as an outcome of the resolution of one or a number of major BM problems, to support other innovation activities. Thus, unintentional BMI processes take off from existing BMs and are characterized by a sequence of major problem formulation and solving which are orchestrated by shifts between experiential and cognitive search for solutions. My observations suggest that the antecedents to BMI may explain why in some cases, BMIs emerge unintentionally and in others firms embark purposefully on BMI. I discuss organizational implementation of BMIs in relation to how firms decide about the degree of separation and integration between parallel BMs. It is argued that the decision about how to structure parallel BMs cannot be made ex ante but emerges through the process of search for a new BM.   The contributions of this thesis are threefold; First it contributes to the emerging conceptualizations of BMI processes by explaining how BMI processes unfold in the two distinct spaces of ‘new BM design’ and ‘existing BM transformation’. Second, the thesis contributes to the BMI literature by introducing problem as a mechanism and theoretical construct for understanding BMI processes in established firms. While the prior literature emphasizes patterns of shift between cognitive search and experiential learning when firms search for a new BM, they do not explain under what circumstances firms embark on either mode of search. Using the problem as the unit of analysis provides an important theoretical basis for conceptualizing the dynamics of the BM by understanding sequential shifts between the two modes of learning along the BMI process. Third, the thesis contributes to the growing debates on how to organize parallel BMs by showing that what is to be separated between the BMs depends on the specific context of the firm. Prior to answering the question of how separated parallel BMs should be, firms need to make sure that they have a viable BM and understand how it operates.