Satellite measurements (retrievals) of surface soil moisture are subject to errors and cannot provide complete space-time coverage. Data assimilation systems merge available retrievals with information from land surface models and antecedent meteorological data, information that is spatio-temporally complete but likewise uncertain. For the design of new satellite missions it is critical to understand just how uncertain retrievals can be and still be useful. Here, we present a synthetic data assimilation experiment that determines the contribution of retrievals to the skill of land assimilation products (soil moisture and evapotranspiration) as a function of retrieval and land model skill. As expected, the skill of the assimilation products increases with the skill of the model and that of the retrievals. The skill of the soil moisture assimilation products always exceeds that of the model acting alone; even retrievals of low quality contribute information to the assimilation product, particularly if model skill is modest.