We study a new market design for K-12 school broadband procurement that switched from school-specific bidding to a system that bundled schools into groups. Using an event study approach, we estimate the program reduced internet prices by 37% per Mbps per month while increasing bandwidth by 500%. These benefits occurred by mitigating exposure risk in broadband procurement -- the risk that providers win too few contracts to cover fixed infrastructure costs. Using a bounds approach, we show robustness of our estimates and document that participants saved at least as much as their federal subsidies and experienced substantial welfare gains.