The integration of artificial intelligence (AI) in next generation cloud-native mobile networks has transformed network and service orchestration, driving the evolution towards fully automated zero-touch operations. As the AI in telecommunications market rapidly expands, the demand for transparency in AI decision-making, known as eXplainable AI (XAI), becomes of paramount importance. XAI mitigates the "black box" nature of AI models, ensuring transparency, crucial for regulatory compliance, user acceptance and system improvement. In this article, motivated by the diversity of data types in cloudnative orchestration scenarios (i.e., tabular, time-series and raw text), we explore the impact that the different data types may have on the selection of the appropriate XAI technique. In addition, we classify and discuss XAI techniques suitable for each data type, outlining a series of use cases in edge/cloud orchestration. Finally, we present two real-world proof-of-concept use cases, demonstrating the benefits that XAI can bring to cloudnative networks. Our contributions aim to facilitate the effective integration of AI in dynamic environments, bridging the gap between AI model interpretability and practical deployment in next generation cloud-native networks.