3 steps to analyze user churn

What is user churn ? This can be judged by the TA indicator, that is, by the two indicators of time (Time) and key behavior (Action), that is, how long the user has not performed the key behavior operation. For example, for a reading product, a user may not have visited the product for more than a month.

1. Determine the nature and type of user churn

To see the essence through data and judge the nature and type of user churn, the most fundamental thing is to judge through data.

1. Determine what kind of users are lost users

What is user churn? This can be judged by the TA indicator, that is, by the two indicators of time (Time) and key behavior (Action), that is, how long the user has not performed the key behavior operation. For example, for a reading product, a user may not have visited the product for more than a month.

In addition, not all user churn can be judged as real user churn. Only by identifying the real lost users can we deal with it appropriately. For example, for internal products of an enterprise, users can only be employees of the company. After the employees leave, they can no longer use the product. This type of user loss cannot be counted as lost users.

2. Characteristics of lost users

Are the lost users new or old users? What needs to be clarified is the boundary between new and old users, which can be determined by the product life cycle. Generally, the determination is as follows:

New Users:

1. Introduction period, contact adaptation
2. Growth stage, exploring growth

Old users:

1. Mature, seeking experience
2. Fatigue period, loss of hidden dangers

Through the product life cycle and the characteristics of the product itself, we can calculate the clear boundaries between new and old users of our products.

3. Behavior analysis before user churn

If new and old users are analyzed horizontally, then the analysis of user behavior before churn is a vertical analysis, that is, the analysis of the time point of churn . Through data analysis, the user's operating behavior before churn and whether certain touch points of the platform or product itself have become the fuse for user churn.

1. Lost users’ own operational behaviors
2. Product/Platform Operation Events

2. Analyze the reasons for user churn

The real reasons for user churn can be found through analysis from two perspectives: product life cycle and pre-churn behavior analysis .

1. Product life cycle

There are different reasons for user churn at different stages.

• A type of new user, who is introduced during contact, immediately quits after contact: the quality of the promotion channel is poor, and the users brought by the channel are not necessarily real and effective users; the product version update leads to...
• The second type of new users are in the process of growth and exploration, and are beginning to use and familiarize themselves with the product: it is difficult to use the product, there is a lack of effective product usage guidance, and the cognitive cost for new users is high; they cannot feel the difference between the product and competing products, and users have preconceived ideas, making it difficult to replace the position of competing products in the minds of users; defects in the product design itself, such as slow product loading, prevent users from fully experiencing the entire business process...
• A type of old users, who are pursuing experience and have used the complete process or core business functions of the product: Generally, this type of users are those whose behavior of installing and using the product is triggered by a passive event, such as downloading the relevant app when applying for personal tax. After completing the operation, the user uninstalls the product...
• The second type of old users are tired of using and become potential lost users: key behaviors are difficult to generate user stickiness, users gradually lose the desire to use, there is a lack of living space for different users, and the product itself has design defects. For example, reading community products have poor quality of posts and comments, which are difficult to bring users the feeling of reading; e-commerce shopping products, users encounter unpleasant experiences during the purchase process, such as slow response from customer service or lack of manual customer service, product delivery timeouts and no platform intervention, etc.

2. Analysis of user behavior before churn

• Lost user's own behavior and operation: Based on the user's operation behavior before loss, it can be judged what kind of operation behavior caused the loss. For example, the user may have filed a complaint during the last operation, or the user may be unable to continue due to a stuck operation process...
• Product/platform operation events: Before losing users, the product/platform causes the loss of users through activities, content and other operations. For example, a product platform that was originally free to use issued a notice that it would start charging, or an app push pushed content at an inappropriate time period. For example, a news app released unconfirmed news in order to seize the opportunity to publish, which caused users’ disgust and displeasure...

3. Develop a churn operation mechanism

The formulation of the loss operation mechanism is mainly based on two points: prevention as the main and recall as the auxiliary. In real life, it is extremely difficult to achieve 100% user recall after user churn, so it is better to do preventive work in advance than to recall users after the fact.

1. Loss prevention mechanism

• Identify lost users: Identify which types of users are considered lost in the true sense, which is also emphasized at the beginning of the article;
• Analyze the signs of churn: Are users actively churned or passively churned? What actions did the platform and product take before users left? For example, for e-commerce products, if users apply for and complain to manual customer service before churn, then such users have signs before churn;
• Establish a churn warning mechanism: Establishing a warning mechanism is mainly to prevent further churn of users. By observing the signs of churned users, we can prevent another group of potential churned users from actually churned. That is, we need to formulate improvement measures. For example, if there are signs that affect the core business process, we can involve product and technical personnel to discuss whether the product system design needs to be changed. If the scope of the impact is small, we can take some soft measures to prevent it.
• Intervention and guidance: For users who show signs of churn, an upgraded intervention mechanism must be established. For example, for users who have filed complaints, you can conduct a return visit through SMS/email/phone calls to listen to their real feedback and answer their confusion. This can alleviate the emotional problems of users before churn and help them solve their problems, leaving them with the impression that they are "valued";

2. Lost customer recall mechanism

(1) Determine whether the user is likely to return

1. Whether users are likely to return: Under what circumstances will users return? For some one-time operation products, the possibility of users returning is extremely low, so the input-output ratio needs to be considered;
2. Whether there are disadvantages to user return: whether the return of users will affect existing core users, and evaluate the impact of user return. For example, some community users habitually spread negative energy and unhealthy content in the community, and are lost due to the system punishment mechanism. For the sake of a healthy community environment, such users can be judged as unnecessary to be saved;

(2) Design of a plan to recall lost users

The design can be done through three ideas, including recall channels, recall methods, and guidance after recall. In addition, there are three key points to note:

1. Increase user personalized operations: Some users may churn simply because the product functions fail to meet the personalized needs of some users. For such users, in addition to feeding back the product through operations as a reference for subsequent product iterations and upgrades, you can also increase user attention through some short-term event operations or content operations.
2. Pay attention to controlling the content and frequency of recalls: Only with appropriate content and frequency can you effectively recall users without making them feel disgusted. Excessive recalls will only increase user dissatisfaction. For example, some apps continue to trigger text messages every day after users uninstall them, which has a counterproductive effect.
3. Add interviews with lost users when necessary: ​​After a series of recall actions, if the user return effect is still not obvious, you can add some one-on-one user interviews and provide some interview incentive measures for users to collect effective user voices.

IV. Conclusion

By judging the nature of user churn, the reasons for user churn, and the user churn operation mechanism, we analyze how to do a good job in user churn prevention. In fact, the key point is still the saying "prevention first, recall second". Don't wait until the user really says goodbye to you before regretting it. Doing a good job of user retention and maintenance is the key. More articles on user retention will be added later.

Author: Internet Xiaohua Miao

Source: Internet Xiaohua Miao