We examine the extent to which the network effects that lead to the adoption of an online social platform are local. Focusing on the fantasy sports market in the US, we find that the size of a county's existing user base on the platform significantly impacts the number of new users who join the platform in that county. However, the size of the user base in nearby counties does not impact adoption, suggesting a local network effect. We also find evidence of heterogeneous network effects, as the impact of the user base is stronger in higher-income counties. Using simulations, we demonstrate that the initial distribution of users across counties can significantly influence the growth of the network over time. We draw implications for optimal seeding based on county population and income.