The recovery of degraded ecosystems is of great concern in our changing world, and the field of restoration is ever growing. However, despite long-term calls for a greater focus on monitoring, both planning and evaluation of restoration success are rarely underpinned by data. Where ecological communities have become depauperate, colonisation processes must be re-instated to facilitate community recovery. Colonisation involves two key stages: dispersal and establishment, both of which are influenced by species traits, local habitat conditions, and landscape context. Here, we introduce a framework to systematically identify and address constraints on the recovery of communities, using macroinvertebrate communities in lotic systems as an illustrative example. The framework incorporates targeted monitoring tools, including drift sampling and aerial adult surveys, to distinguish dispersal from establishment limitations and inform appropriate interventions. By systematically assessing recovery constraints, the framework can be used to identify suitable restoration actions. For example, measures to facilitate dispersal (such as riparian planting or community translocations), or measures to support establishment (such as improving water quality and habitat availability). In practice, monitoring is often insufficient to identify specific limiting factors, with practitioners resorting to off-the-shelf restoration methods which may not be appropriate. However, the tools to comprehensively assess communities in lotic systems do exist; we need to improve understanding and accessibility amongst practitioners to champion their use in restoration. Plans underpinned by data will include more strategic restoration methods, thereby improving restoration success.