Full Paper - Regular Research Paper: Workarounds are goal-driven adaptations of business processes that employees implement to overcome perceived constraints at work. While process deviations can be easily detected with data-driven methods like process mining, it is hard to distinguish workarounds from related, yet distinct, concepts. The SWORD framework constitutes a state-of-the-art method for the data-driven detection of workarounds in business process event logs based on pre-defined patterns extracted from support processes in the healthcare domain. However, currently, SWORD has solely been applied to highly standardized processes with low variation and knowledge intensity, while it can be assumed that workarounds more often appear in the latter and also bear bigger potential for innovating processes. Furthermore, SWORD neither comprises data preparation steps nor enables the analysis of workarounds and their implications for the organization. In this paper, we develop an exaptation of the existing SWORD framework, coined LongSWORD, together with two industrial case organizations. Our contribution to theory and practice is threefold. First, we present a framework that enables the preparation of a meaningful event log in alignment with according process model or routine. Second, the detection and analysis of workarounds in core industrial processes is enabled by adding two new cross-case patterns. Third, the LongSWORD method enables others to assess the implications of workarounds beyond its individual implementers.