Learn How to Experiment: The Development of a Corporate Accelerator

Incumbent high technology organizations’ responses when facing digital disruption present an important area for research in the field of strategy and innovation management. In times of increasing environmental dynamism and uncertainty, these firms, which previously have relied on strong capabilities in research and development, supported by experiential learning, find themselves unable to learn about new markets and technologies in a satisfactory pace. This study argues that these firms must learn to experiment in order to facilitate development of new businesses, thus enabling organizational renewal and adaptation to the changing environment. As little is known about how incumbents may achieve this undertaking, this thesis sets out to answer the following question: How do large established firms learn how to experiment in order to develop new businesses? Through an in-depth case study of a large multinational telecommunications company, I follow the establishment of a corporate accelerator aimed to support systematic experimentation- driven venturing in new strategic business areas. The study draws on multiple sources of data, including field observations, interviews and archival data. The findings, which amount to a grounded model depicting the process of learning how to experiment, are discussed in the light of corporate entrepreneurship theory, the emerging literature on accelerators as enablers of experimentational learning, and organizational learning literature. I identify six distinct learning mechanisms that transpire partly in parallel, partly sequentially over three distinct phases. These learning mechanisms are referred to as (1) schematic arms- length vicarious learning; (2) coactive vicarious learning; (3) unlearning of conflicting corporate practices; (4) replication across the organization; (5) deliberate learning; and (6) adaptive replication of corporate practices.