With the fast growth of e-commerce and the emerging trend of “New Retail”—that is, online and offline integration—the important research issues are how to know the best ways to collect and analyze users’ search behaviors online for a streamlined shopping process. Accordingly, we proposed a search pattern analytical method to analyze users’ search behavior in the entire shopping process on the target website from the perspective of the users’ need states. We have focused on the recommendation functions (RFs) and the search functions on Taobao.com to evaluate the effectiveness of each RF to support the online shopping process in different user-need states, namely in a goal-oriented or an exploratory-based approach to online shopping. We first adopted zero-order state transition matrices and then used lag sequential analysis (LSA) to derive the significant repeating search patterns. The results show that the goal-oriented shoppers tend to search directly, whereas exploratory shoppers tend to explore the categories of products as their initial RFs. In addition, goal-oriented users have much more simple search paths compared to the exploratory-based users when engaged in online shopping. Furthermore, based on the results of the LSA, there are two typical search patterns for goal-oriented users and no search pattern for the exploratory ones. Interestingly, the results reveal that exploratory-based users are easily stimulated by context even if they have moved to specific stores. The aim of this research is to summarize users’ search paths and patterns with different need states to help the e-store design the website.