User stratification is very important to product development. By stratifying users, we can better understand the needs, behaviors, and preferences of different user groups, thereby positioning and designing products more accurately, which is conducive to improving market competitiveness.
Without user stratification, product managers may not understand user needs, be unable to provide personalized products and satisfy user experience, and find it difficult to accurately position products and respond to market competition, which is not conducive to the realization of product value.
The common user stratification mainly has the following four methods:
1. User value segmentation and stratification
User value segmentation and stratification refers to judging the value of user contribution through a certain dimension of the product. There are two common layering methods: segmenting based on user life cycle and segmenting based on key user behavior.
1. User life cycle model
The definition of user life cycle must be related to the growth path of user value, and different types of products have different value growth paths. According to whether the product is paid or not and the frequency of use of the product, the value growth path of different products can be divided into four quadrants:
The general user life cycle is divided into 5 stages:
Introduction period : the user has just registered, has not yet experienced the core function process, and is not familiar with the value brought by the product;
Growth period : have experienced the core functional process, are more recognized for the value provided by the product, and have established initial usage habits;
Maturity period : A high degree of usage habit and dependence on the product is formed, the frequency and duration of use are significantly higher, and it can contribute higher value;
Dormant period : Once a mature user, but now no longer visits and uses the product or the frequency of visits is getting lower and lower. If not used for more than 10 days.
Lost period : the product has not been logged in for a long time, and the product has even been uninstalled. Such as users who have not used it for more than 30 days.
2. RFM method
The RFM method is a typical and commonly used method for segmenting and stratifying user values based on key user behaviors. RFM represents three key behaviors of users:
R (Recency): distance from the last transaction
F (Frequency): transaction frequency
M (Monetary): transaction amount
Users can be divided into 8 types through the RFM method:
The core logic of the RFM method is to find out the key behaviors that affect the value of users, and then conduct cross-analysis and user segmentation.
2. Segmentation and stratification of user needs
User needs are segmented and layered, mainly divided into two dimensions: user natural attributes and user personalized needs.
User natural attributes : mainly rely on the user's basic data, including gender, age, occupation, income, etc.;
User personalized needs : mainly rely on user behavior data, personal consumption preferences, and personal scene preferences.
Segmentation and stratification of user needs is mainly by analyzing whether users are in these two dimensions, and there are obvious differences in user needs. The method of judgment: experience insight and data speaking. In specific operations, we can use a single dimension for distinction, or use two dimensions for cross analysis.
In order to further improve the efficiency and quality of demand analysis, we can use development tools to quickly improve the efficiency of demand analysis. For example, the CoCode development cloud uses GPT technology to automatically generate standard user stories from user requirements through requirement entry and automatic decomposition of sub-requirements ; the requirements analysis tool uses AI to pass requirements testing and consistency testing, and can quickly analyze within a few minutes Defects in user requirements, such as ambiguities, repetitions, omissions, inconsistencies, and complexity, etc., accurately pinpoint requirements problems, thereby modifying requirements defects more efficiently and improving the quality of user requirements analysis.
3. Layering of AARRR model
AARRR model stratification can be used not only for growth, but also for user stratification. It includes 5 links:
User acquisition (Acquisition): Users register products from different channels, they need to pay attention to the registration conversion rate of different channels, and optimize channel resource allocation.
Activation : The user has registered, but has not completed the core process experience of the product. At this stage, it is necessary to guide users to complete the core process.
Improve the retention rate (Retention): Experience the core process, but the retention time is not high, analyze this retention problem, and give a specific operation strategy.
Viral transmission (Referral): For users whose active frequency exceeds a certain range, we can use tools to optimize and stimulate users to spread.
Revenue : For users whose activity and retention time exceed a certain range, we can strengthen payment guidance for specific users and in combination with specific scenarios.
The layering of the AARRR model is suitable for the relatively primary stage of the product. At this stage, the number of users is average, and the company's data system has not yet been established.
4. User Identity Segmentation and Layering
Generally, in the content community, users are divided into KOL and ordinary users by user identity, and the operation strategies for these two types of users are definitely different. In the field of the product, when the user's behavioral characteristics and appeals are quite different, it is more appropriate to use identity segmentation for stratification. For example, Weibo divides users into: star users, KOL users, active users and ordinary users.
If there is no relationship between users, then this model is not applicable; if there is a relationship, because of the contribution or scarcity of the user class, then the user hierarchy model can be built according to the contribution and scarcity; if users of different stages It can be advanced naturally, so build a user hierarchical model based on the advanced class.
The purpose of our user stratification is to refine operations for users, so as to maximize product benefits with minimum operating costs. Its core is: users at different levels can be defined with clear data labels and attribute labels, so as to realize the automation of user labeling; and user operation strategies at different levels are targeted and stable.