Recently, in the special live broadcast of "Everyone is a growth officer of data-driven APP users' refined operation", an Internet solution expert solved the "incurable diseases" of APP users' operation online, which not only brought us the construction method of closed-loop system of APP users' operation, the effective strategy of users' life cycle operation, but also exclusively analyzed the user growth case of industry-leading APP. The following is the essence of this live broadcast. Developers and operators should check the construction method of APP user operation system. At the moment when the traffic dividend is gradually peaking, the operational focus of APP is gradually shifting from Racine to the refined operation stage of existing users. Building a data-driven user refined operation system has become a new engine for the continuous growth of APP. However, in the process of exploring users' refined operation, APP also faces a series of difficulties and challenges: 1. How to get through multi-terminal data such as APP, web, H5 and applet, and make full use of data assets? 2.2 How? APP supplements external tag data, improves user portraits, and reduces the cost of tag system construction? 3. How to use a platform to complete user insight, stratification, circle selection and reach, and improve user operation efficiency? In view of this series of problems, the user operation platform is launched to help APP build a perfect user operation system from three aspects: user analysis, user insight and user access: First, it has a solid data foundation and effectively integrates the user behavior characteristic data of the buried point statistics to form a complete user data platform, which is convenient for subsequent data analysis and operation; The second is to have a portrait insight into the user behavior characteristics, and label the user behavior characteristics by combining the APP's own tags and personal push data tags; Third, docking with its own message push platform, APP can generate user tags or user groups by using the results of user insight, and automatically synchronize the created user groups to a push message push platform, so as to achieve one-click access to users, efficiently connect users, and achieve refined hierarchical access. Analysis of Typical Scenarios of User Refined Operation What are the effective strategies for APP reuse in different scenarios such as user tag system construction, user stratification, activity operation and content operation? Next, we disassemble them one by one. 1. Building a user label system and understanding users is the first step of user operation. For this reason, APP needs to build a perfect user tag system to identify user portraits in different scenarios, so as to provide users with differentiated content and services. So, how to build a label system? APP can be done by pushing user operations. First of all, pushing user operations can effectively enrich APP data tags. In addition to the basic attributes such as the user's age, occupation and consumption level, as well as the user behavior characteristic data such as the user's visit, consumption, collection and purchase concern at the APP's own terminal, the APP can also supplement the data outside the APP through a push data tag to fully understand the user's interests and offline scene preferences. At the same time, personal push user operation provides a label management platform, supports APP customization to create labels, labels its own internal data through statistics of buried points, or combines basic labels, industry labels and thematic labels precipitated by pushing big data to build a user label system efficiently. 2. User Hierarchical Operation In addition to in-depth user analysis, APP can also do refined user stratification of individual business based on the tag system, and operate around the whole life cycle of new users, such as cold start, silent user promotion, and lost user identification. , improve user retention and conversion. In the cold start stage of new users, with the help of push users, APP can quickly understand the attributes and interest preferences of newly registered users, complete user portraits, and recommend matching content for users when they contact the first contact of APP. For active users, the APP can circle the silent users through the active data in the terminal and analyze the portraits of the silent users. At the same time, the value of silent users can be scored according to their historical activity and consumption, and the value of silent users can be effectively evaluated. In the aspect of lost user identification, pushing users to operate products can help APP effectively distinguish silent users from lost users, find the high TGI (significant characteristics) of lost user groups, and quickly evaluate the recovery cost. For example, on the personal push user operation platform, APP operators can effectively identify the historical behavior characteristics in the lost user terminals by browsing the live content of e-sports through TGI- Love, thus increasing the content recommendation of e-sports events in a targeted manner, greatly increasing the chances of users' redemption. 3. Activity Operation Activity operation is becoming more and more important in the APP operation system. In the face of major festivals, such as 6 18 Zhong Da promotion, Double Eleven, year-end promotion, New Year Festival, etc., APP needs to do some activities to improve user activity and purchase conversion rate. How to do a good job in activity operation? To sum up, it can be divided into four stages: activity data statistics, activity user analysis, activity plan formulation and activity recovery. First of all, the APP can restore user behavior links and gain insight into the traffic value of various activity resources, including search box, carousel, classified navigation bar and so on. Through the statistical analysis function of multi-end buried points operated by users. Secondly, in different scenarios such as activity drainage, user promotion and user transformation, operators can flexibly use analysis models such as event analysis, funnel analysis and path analysis to analyze the transformation of activities in all directions, build customized analysis kanban, realize the visualization of activity data and analyze the effect of activities. For example, the e-commerce app can analyze the conversion rate of key nodes in the activity path such as activity banner click-activity detail page browsing-activity participation-coupon collection through the funnel analysis function, and find the growth point of GMV. In addition, through the user running activity planning module, operators can directly select all the target groups of this activity, configure the activity objectives, and arrive at the activity welfare point regularly, repeatedly and differentially before the user completes the objectives, so as to automatically stimulate users and achieve the operation objectives efficiently. Finally, in terms of activity recovery, APP can conduct insight analysis on the activity crowd, and fully recover the activity data effect and continuously optimize the operation effect by understanding the results of key indicators such as the total number of participants, the number of trading users, the activity conversion rate, real-time PV and UV. 4. Content operation Content operation is the basis for supporting activities, products and user operations. Personal push user operation helps APP to build an intelligent content recommendation mechanism by linking content tags with user behavior and tag system. APP can reasonably allocate resources such as Banner, list page and search box based on user behavior characteristics, reading habits, search preferences and interest preferences, so that users can obtain interested content more efficiently. At the same time, the complete content tag and user data tag system also provides data for prediction model training for the client intelligent recommendation system. Through data tag supplement, it helps APP to iterate the content recommendation model continuously and improve the efficiency and effect of content recommendation. 5. User Access By pushing the user operation platform, APP can not only conduct user insight and analysis efficiently, but also push its own message push platform to reach and connect users with one click, thus improving user activity and content conversion. With the user group management function operated by Twitter users, APP can create user groups through user tags, user attributes, event behaviors, etc., and automatically synchronize the created user groups to the Twitter message push platform to achieve hierarchical and refined message reach. APP can also push the smart copy library through a push message, quickly find the matching relationship between the crowd and the copy, automatically configure the high-click copy of the target crowd, and realize the smart message push. For example, an e-commerce APP selects price-sensitive users in the promotion of 6 18 through the user operation circle, and the system automatically configures multiple keywords such as "discount" and "price reduction". Through the application of high-click copybooks, the conversion rate of commodities has been greatly improved. At the same time, the personal push user operation platform can also assist the APP to output the copywriting preference characteristics of segmented people, and provide editing suggestions to help the APP maximize the click-through rate of message push. At present, personal push user-operated products have reached cooperation with more and more head customers, which are widely used in e-commerce, news information, tourism, audio-visual and games industries. Next, we will combine the practice of an information APP digital operation scenario to provide a successful model for the refined operation of data-driven APP users. An information APP of a newspaper group has set up its own APP, WeChat applet, WeChat official account, Weibo and other multi-channel content media channels. With the help of push-type user operation platform, APP realizes multi-end data burying statistics, integrates user behavior characteristic data in various channels such as searching, browsing and clicking preferences, and builds a user data platform that can meet different business scenarios and be accessed at any time. At the same time, on the tag management platform provided by Twitter users, APP not only tags its own internal data, but also builds a rich portrait tag system by adding big data tags, creating custom tags, importing its own data tags, and reusing industry tag templates. To achieve efficient user insight and analysis. Through the cooperation with Twitter, the APP has opened up the operational link of data analysis-portrait insight-user access, achieving a high growth rate of 7% and 24% for new users the next day. The sharing of the above live content hopes to provide ideas, methods and experiences for developers to operate the APP in a refined way. "Everyone is a growth officer" is an operation-oriented and product-oriented brand special event, which aims to bring you the cutting-edge knowledge of operation and growth, help you establish the thinking of data growth, build a user's life-cycle operation system, and realize data-driven business growth. Everyone is a growth officer, and the next issue is even more exciting. Welcome everyone to continue to pay attention!