Learning Analytics Dashboards (LADs) are increasingly used to support data-informed teaching and to improve student outcomes. Although research in this area has grown significantly, questions remain on how these dashboards are designed, implemented, and used in practice. This study delivers an umbrella review of 17 systematic reviews published between 2015 and 2025, bringing together existing evidence on learning analytics dashboards. The analysis looks at three main aspects: the elements included in dashboards, the technologies and tools used to develop them, and their impact on teaching practices.

Focusing on teachers’ practices, the data show that dashboard research is gradually moving from a purely technical emphasis to more pedagogically informed approaches aimed at supporting teachers’ decision-making and helping them better comprehend student learning processes. However, multiple challenges still remain. While dashboards are effective in raising awareness, they frequently fall short in supporting concrete actions. Teachers are often involved in initial design stages but less so in later development and evaluation, and many studies rely on short-term or self-reported measures. This may explain why most dashboards remain at the prototype stage. In addition, while artificial intelligence is often used for prediction, limited attention is given to transparency or instructional use. Issues such as privacy and bias also receive relatively little consideration.