We first investigate the coverage of location information in social networks. Intuitively, the location information in SNS would contain a subset of the places in real life, as social media captures a relatively small fraction of our lives by people who are active participants of SNS. To estimate the coverage, we manually matched the places visited by users in daily life to the locations in SNS. We used the labeled names and offline feedback from the participants. Figure 1(a) shows that SNS misses about 22% of the places users visited. The missing places are mostly food or shopping places with low popularity, or recently built places, as shown in Figure 1(b). Considering that our dataset is collected with 70 smartphone users, the portion of missing places would increase with more participants since the manual registration of places in SNS cannot contain all the places in real life. Another finding is that mismatch exists in user’s understanding of the actual place and the locations provided by SNS. For example, a user stays at the NexOne Inc. office, but SNS provides this place as the name of the building. A user would think that the building name could not express the place she visited, since the building has many offices, stores, and restaurants. Figure 1(a) shows that the mismatch ratio is about 14.5% of the places in our dataset. The mismatched places were mostly located in college and university regions, shopping districts, or office buildings, as shown in Figure 1(c). In summary, our preliminary analysis indicates that crowdsensing using smartphones has the potential to expand the coverage and improve the granularity of location information that are missed by social media.