Blanket advice on optimal fertilizer application rates has failed to achieve potential yield gains for crop production in much of Sub-Saharan Africa. However, digital technology now makes it possible to deliver personalized extension services to farmers at a much lower cost. In this paper, we present results from a randomized control trial designed to evaluate the effectiveness of a mobile app that provides personalized advice on rice nutrient management. The experiment induced variation in both access to the app and access to an input grant for fertilizer. We find that households with access to the app have 15 percent higher yields and 20 percent higher profits. We show that, on average, personalized advice increases yields without increasing the overall quantity of fertilizer used. Rather, some households increase their use of fertilizer while other households decrease their use of the input. We conclude that the scaling of personalized extension services could improve productivity and livelihoods in Sub-Saharan Africa without necessarily increasing the total amount of fertilizer in use.