Drawing on studies of strategic dynamics and organizational change, this thesis proposes four approaches to balancing exploration and exploitation over time: Specialist, Cyclical, Irregular, and Regular. Various approaches to ambidexterity may be more effective under different environment conditions, and performance may vary along with: 1) varying types of rule change environments, 2) varying levels of competitive intensity among firms, 3) reactive versus proactive timing heuristics, and 4) varying levels of product diversification. Several hypotheses are developed and confirmed using qualitative field research and agent-based modeling. Results indicated that strategic leaders should balance their exploration and exploitation with Regular ambidexterity as their environments become dominated by competence enhancing innovation. Conversely, firms should temporally shift their balance of exploration and exploitation when competence-destroying changes dominate. In a balanced environment, Irregular ambidexterity performs best. These finding are especially relevant in highly competitive contexts. Also, proactive switching increases performance more than reactive switching, whereas diversification reduces the performance of sequential heuristics.

Drawing on studies of strategic dynamics and organizational change, this thesis proposes four approaches to balancing exploration and exploitation over time: Specialist, Cyclical, Irregular, and Regular. Various approaches to ambidexterity may be more effective under different environment conditions, and performance may vary along with: 1) varying types of rule change environments, 2) varying levels of competitive intensity among firms, 3) reactive versus proactive timing heuristics, and 4) varying levels of product diversification. Several hypotheses are developed and confirmed using qualitative field research and agent-based modeling. Results indicated that strategic leaders should balance their exploration and exploitation with Regular ambidexterity as their environments become dominated by competence enhancing innovation. Conversely, firms should temporally shift their balance of exploration and exploitation when competence-destroying changes dominate. In a balanced environment, Irregular ambidexterity performs best. These finding are especially relevant in highly competitive contexts. Also, proactive switching increases performance more than reactive switching, whereas diversification reduces the performance of sequential heuristics.

For robotic systems to continue to move towards ubiquity, robots need to be more autonomous. More autonomy dictates that robots need to be able to make better decisions. Control theory and machine learning are fields of robotics that focus on the decision making process. However, each of these fields implements decision making at different levels of abstraction and at different time scales. Control theory defines low-level decisions at high rates, while machine learning defines high-level decision at low rates. The objective of this research is to integrate tools from both machine leaning and control theory to solve higher dimensional, complex problems, and to optimize the decision making process. Throughout this research, multiple algorithms were created that use concepts from both control theory and machine learning, which provide new tools for robots to make better decisions. One algorithm enables a robot to learn how to optimally explore an unknown space, and autonomously decide when to explore for new information or exploit its current information. Another algorithm enables a robot to learn how to locomote with complex dynamics. These algorithms are evaluated both in simulation and on real robots. The results and analysis of these experiments are presented, which demonstrate the utility of the algorithms introduced in this work. Additionally, a new notion of “learnability” is introduced to define and determine when a given dynamical system has the ability to gain knowledge to optimize a given objective function.

Organisational ambidexterity is considered a key to company survival and performance. Despite this, organisational ambidexterity is still a poorly understood phenomenon, especially in an SME context. The purpose of this study was therefore to investigate how the compliance with ambidextrous elements is perceived at different levels in manufacturing SMEs, to increase the understanding of organisational ambidexterity in this context. The empirical data was collected through a combination of questionnaire and interview. The case companies in this report perceive that they comply stronger with contextual elements than with structural elements. The strong compliance with contextual elements is motivated by the lack of hierarchies, flexibility in the company, different management structure and low number of employees. This allows employees to perform the contextual elements such as initiative-taking, cooperating, brokering and multitasking. The structural elements including e.g. vision, values, strategies, senior team responsibility and alignment are perceived differently at different  hierarchal levels, indicating that there are subcultures within the hierarchal levels within a company. The biggest difference can be found between the middle managers and the top managers,. Workers perceive that they are not included in explorationb within the company, and that the exploration occur more sporadically than those for exploitation. The definitions of exploration and exploitation vary between the companies which results in a lack of consensus. This makes it difficult for the companies to perform the changes necessary in order to develop and achieve long-term sustainable growth i.e. economical sustainability. The managerial implication of this report concerns four actions: (1) create a common definition for exploration, (2) develop goals for exploration, (3) communicate for buy-in and (4) involve all employees.

<p>In this thesis we address the problem of how to construct an optimal algorithm for displaying banners (i.e advertisements shown on web sites). The optimization is based on the revenue each banner generates, with the aim of selecting those banners which maximize future total revenue. Banner optimality is of major importance in the e-commerce industry, in particular on web sites with heavy traffic. The 'micropayments' from showing banners add up to substantial profits due to the large volumes involved. We provide a broad, up-to-date and primarily theoretical treatment of this global optimization problem. Through a synthesis of mathematical modeling, statistical methodology and computer science we construct a stochastic 'planning algorithm'. The superiority of our algorithm is based on empirical analysis conducted by us on real internet-data at TradeDoubler AB, as well as test-results on a selection of stylized data-sets. The algorithm is flexible and adapts well to new environments.</p>

Background: Responding to fast technological and environmental changes brings about challenges and paradoxes for companies that should be resolved in order to survive long-term and to achieve a sustainable competitive advantage. Ambidexterity is considered a solution to organizational paradoxes. Aim: The purpose of this research is to explore how firms can achieve ambidexterity to handle organizational paradoxes in different market conditions using their dynamic capabilities. Definitions: Ambidexterity: The ability of organizations to handle adaptability and alignment, exploration and exploitation at the same time Dynamic capabilities: The firm’s ability to integrate, build and reconfigure internal and external competencies to address rapidly changing environments Exploration: Activities such as innovation, discovering new opportunities, variation Exploitation: Activities that concern efficiency, implementation and execution Results: There are possible options that companies can follow to achieveambidexterity. These sets of options are distinguished as external vs.internal, sequential vs. parallel, structural vs. contextual and the role ofsenior management behaviour. Depending on market dynamism andenvironmental conditions, a different set of options could be suitable fordifferent companies. In addition, companies can enhance the likelihoodof achieving ambidexterity using their dynamic capabilities.

In this thesis we address the problem of how to construct an optimal algorithm for displaying banners (i.e advertisements shown on web sites). The optimization is based on the revenue each banner generates, with the aim of selecting those banners which maximize future total revenue. Banner optimality is of major importance in the e-commerce industry, in particular on web sites with heavy traffic. The 'micropayments' from showing banners add up to substantial profits due to the large volumes involved. We provide a broad, up-to-date and primarily theoretical treatment of this global optimization problem. Through a synthesis of mathematical modeling, statistical methodology and computer science we construct a stochastic 'planning algorithm'. The superiority of our algorithm is based on empirical analysis conducted by us on real internet-data at TradeDoubler AB, as well as test-results on a selection of stylized data-sets. The algorithm is flexible and adapts well to new environments.

Consumers make trade-offs in their purchase decision making between extending market knowledge from exploring a product market and maximizing purchase value based on exploiting their current knowledge. The value of these strategies can be enhanced opportunistically by taking advantage of promotions. In this research, a new and unique datamining model was developed to process store scanner data for quantifying the brand selection behaviour of an individual consumer in reaction to promotions. Selected consumers in each of Pittsfield’s salty snack, yogurt, and toilet tissue markets were then segmented into four behavioural segments using clustering analysis based on their Prevalence of Promotion and their Value of Information from Purchases. The behavioural segmentation was valid, as the four generated behavioural segments in each product market could be differentiated by using demographic variables. In a product market, the demographic profiles of behavioural segments can be generated andused for improving the performance in targeting consumers. The generated demographic profiles of a behavioural segment explain how consumers in the segment react to promotions,which can be used for predicting how consumers will react in the future. Complementing demographic profiles, dynamic behavioural evolvements in consumer purchase lifecycles can also help to predict the purchase behaviours of consumers in the future from the purchases that the consumers have made. The evolvements enable people to understand how consumers with a given amount of market experiences make their purchase decisions via making trade-offs between market knowledge extension and immediate purchase value maximization. Product markets differ in their available number of brands for selection. The findings generated in a product market, however, cannot be generalized to a different product market. Consumers have different demographic-related purchase behaviours across frequently purchased consumer goods markets. Based on the findings in the research, the dissertation discusses and provides suggestions forretail businesses to improve their performances for achieving a competitive edge.

Background: Companies in their search for value creation face different challenges when undertaking digital transformation (DT). Incumbent firms struggle differently than smaller-sized or newer firms because of their characteristics. Therefore, there is a growing interest in defining what types of challenges incumbents face, and understanding what the viable strategies are, necessary to implement digital strategies efficiently. For instance, knowledge exploration and exploitation are key to DT, affecting the recombination of knowledge in different degrees, therefore, also impacting the reach and efficacy of DT itself. Hence, the understanding of knowledge creation, its sharing, and application in the setting of incumbent firms must be improved, providing a better understanding of how digital technologies can be effectively implemented within these types of firms.  Purpose: The purpose of this study is to characterise the challenges that incumbents face during their DT journey, as well as to provide suggestions of strategies to adapt to them. This work also illustrates the role and importance of knowledge exploration and exploitation, and recombination throughout the firm. The goal of the study is to provide a framework for undertaking DT successfully.  Method: The present study applied an inductive qualitative research approach with a single case study research design. The case study company was Telia Infra and included interviews with ten managers from all its different departments. The interviews were of a semi-structured nature to allow flexibility in the interview design.   Conclusion: The highlights of the study’s outcomes are: (i) the disclosure of a relationship between barriers during DT and the digital maturity of the department, showing that knowledge sharing activities are beneficial in the longer term. (ii) A human-centric driven DT strategy, balancing explorative and exploitative activities is incredibly positive for the DT journey. This drives for (iii) exponential change strategies which results in more favourable environments for overcoming barriers and enabling DT.

Using data from the biopharmaceutical industry, this study analyzes the relationship between internal and external exploration (exploitation) and the moderating role of firm characteristics. We argue that the characteristics of a firm - in particular a firm's vertical scope - are an important but so far overlooked contingency factor determining the relationship between internal and external exploration (exploitation). Our findings indicate that differences in vertical scope result in alternative combinations of internal and external exploration. While specialized firms tend to focus either on exploration or exploitation both within the firm as well as with their alliance portfolio, vertically integrated firms prefer to balance exploration and exploitation across firm boundaries. (authors' abstract) / Series: Working Papers / Institute for Strategic Management