PurposeIn this study, the authors use the actor-network theory (ANT) as a theoretical framework to better understand constructing learning as part of the networking process to produce innovations. Focussing on the antecedents of innovation within three teams in an engineering company, the authors propose a framework to enhance understanding of the innovative processes. The authors apply ANT to examine how informal learning is distributed amongst human and non-human actors.Design/methodology/approachBased on 27 interviews in a large Australian engineering company, the authors' qualitative investigation shows that innovation can have very different antecedents. The authors mobilised ANT as the authors' vantage point to explore inanimate actors and their effect on social processes or, more specifically, networks and informal learning.FindingsThe authors propose a framework to better understand innovative processes by exploring the network aspects of non-human actors and their connection to learning. More specifically, findings contribute towards a more granulated understanding of how networks, learning and non-human actors contribute towards innovations in organisations.Practical implicationsThis study has three significant implications for managers and organisations looking to improve their innovation processes. Firstly, fostering open communication is essential for developing successful innovation processes. Secondly, a close relationship with the customer and/or the final users has often been found to positively contribute to innovation processes. Finally, intrateam motivation is also critical when it comes to creating an environment that supports innovation processes.Originality/valueSurprisingly, leadership, communication and motivation did not give the best innovative outcome as the authors expected. Challenging traditional theorisations, low teamwork spirit and high individual performance orientation were some of the powerful drivers of highly innovative teams.