Online Social Networks (OSNs) have become an efficient media to disseminate content like news and advertisements. In this paper, we are interested in investigating the open problem of how does the content popularity influence the information coverage exactly. Specifically, we assume that each user has q friends and forwards the received content to each of his friends with the forwarding probability p, where p is only determined by the content popularity. Based on a geography-aware social model leveraging the forwarding behavior of OSNs users, we explore the exact relationship between the content popularity and the information coverage by the method of network partitions. Then, we mainly have the following observations: 1) The information coverage is only related to p and q; 2) In order to disseminate the content to a constant portion of all users within the network, the forwarding ( 1 ) probability p should be no less than Θ(1/q). Extensive evaluations q on both the generated data and the real OSN topology validate our observations.