Corpora, large electronic collections of texts, have been used in language teaching for several decades. Also known as Data-Driven Learning (DDL), this method has been gaining popularity because empirical research has consistently shown its effectiveness for learning. However, corpora are still underutilized, especially with learners of languages other than English, at lower proficiency levels, and in non-university contexts. This is regrettable because DDL has a great potential for developing modular flipped content, especially for hybrid, remote, and online courses. This article first provides an overview of DDL applications and findings of empirical research. Next, it outlines obstacles to wider DDL implementation as well as available and possible solutions. Corpus user guides and exercise collections tied to specific corpora are discussed as one promising direction, and an example of such new open educational resources for teaching German is presented. The article concludes with a discussion of implications and future directions.