Web log is a pouch of valuable information that records users search queries and related actions on the internet. By mining the recorded information, it is possible to exploit the users underlying goals, interests and search behaviors. In order to mine information from web logs, the web logs should be segmented into sessions or tasks by clustering the queries. In this work, Task Trail is introduced to understand user search behaviors. A Task can be defined as set of semantically relevant queries issued to satisfy an atomic user information need. A task trail represents all user activities within the particular task, such as query reformulations, URL clicks. In most of the previous works, web search logs have been studied mainly at session or query level where users may submit several queries within one task and handle several tasks within one session. Although previous studies have addressed the problem of task identification, little is known about the advantage of using task over session or query for search applications. Instead of analyzing Session Trails or Query Trails, Task Trails can be analysed to determine the user search behaviour much more efficiently. By separating different task trails from a session, it can be used in several search applications such as determining user satisfaction, predicting user search interests, and suggesting related queries.