Learning ambidexterity as ecosystem capability

Organizations are increasingly facing the challenge of learning from what they know while also adapting to what they have yet to discover. Although organizational ambidexterity has long been studied, existing research offers limited understanding of how organizations learn ambidexterity across ecosystems.This paper develops a multi-phase conceptual framework of learning ambidexterity as an ecosystem capability. We theorize learning ambidexterity as a distributed capacity through which organizations selectively engage proximity-based ecosystems to use and create knowledge across the learning phases of exploitation and exploration. Cumulative learning in local ecosystems enables incremental exploitation through internalization and combination. Recombinative learning in proximal ecosystems supports the exploration of known alternatives through boundary-crossing externalization and combination. Generative learning in distant ecosystems enables the exploration of unknown alternatives through socialization and externalization.The study contributes to the organizational learning literature by reframing ambidexterity as recursive, non-linear learning phases that evolve across diverse knowledge environments.