Over the past two decades, online social networks have attracted a great deal of attention from researchers. However, before one can gain insight into the behavior or structure of a network, one must first collect appropriate data. Data collection poses several challenges, such as API or bandwidth limits, which require the data collector to carefully consider which queries to make. Many network crawling methods have been proposed; however, their performance depends on network structure. In particular, our previous work in  has shown that existing algorithms tend to either (1) Do well at exploring dense areas of a network, but have difficulty in transitioning to new areas of the network, or (2) Easily move between network regions, but fail to fully explore each region. In this work, we introduce DE-Crawler, a novel network crawler that attempts to capture the best of both worlds. DE-Crawler consists of two main stages: Densification, in which the crawler aims to find as many nodes as possible in the current dense region (or community), and Expansion, in which the crawler tries to escape from its current region and move to another dense region. We show that DE-Crawler performs well across networks with different structural properties, outperforming baseline algorithms by up to 28%.