The fundamental purpose of user operation is to maximize customer value, fully understand the user's state through data, and then give targeted operation strategies. How to use data to drive user value growth? Sort out the operation process, collect user data, process user data, establish user portraits, accurately reach data feedback, adjust and sort out the operation process, and clarify the common operation purpose. The user operation process has different stages, corresponding to new users, active users, registered applet users, paying users and losing users. The operation can adopt different operation methods according to the links where different users are located. User data collection and collation User basic data field: User's social information data, such as name, gender, date of birth, native place, marriage, education, mobile phone, email address, etc., are mainly filled in by users. User behavior data field: data record of user's operation behavior on products, such as reading content, like, commenting, sharing, etc. User behavior data reflects the user's platform behavior, which is mainly obtained through technology as the data embedding point. User data processing method User hierarchical data analysis: hierarchical classification of users according to certain logic. Central idea: hierarchical classification of users according to core business processes. Example: Analyze users hierarchically according to the core business process shown in the above figure. As a supplement to user stratification, users are grouped at the same level to achieve more refined operations. Users need to be further grouped on the basis of user stratification. For example, the grouping of paying users in the pet community is based on the consumption amount. 2. Grouping based on consumer goods. 3. Grouping based on users' gender. After cross-analysis and grouping, it can be clearly seen that women's consumption intention is greater than that of men, and dog users' consumption intention is greater than that of cat users. RFM user value data analysis: latest consumption date, consumption frequency, consumption amount, latest consumption time, consumption frequency, consumption amount, MonetaryRFM matrix diagram, user loyalty data analysis L is the user loyalty index, N is the time period, T is the time window, and S represents the consumption times, representing the consumption times in a certain period from now. The calculation time window of user loyalty should be combined with its own actual business scene, and the time factor is usually considered in the actual use process, that is, the closer the purchase time is, the higher the user loyalty is. Correctly build user portraits and supplement user attributes as much as possible, such as hobbies, registration time, last speech time, last consumption time, etc. Analysis and summary of user operation data: according to the setting of operation purpose, sort out the operation process and sort out the key behaviors that users need to do. 2. The more user data is collected, the more dimensions can be analyzed. 3. User data processing User stratification, user grouping, RFM user value analysis, user loyalty analysis, the most commonly used user data processing methods 4. The established user portrait can only be included in the user portrait field, and different products have different user portrait fields 5. After users are familiar with the status of each user (user portrait), they can accurately reach 6 through push, email, SMS, Banner and other channels. The data feedback adjustment continuously adjusts the user data processing model and the user portrait field according to the user guidance effect.