Operational Tips: My experience in user operations at Tencent is all here

At the beginning of the new year, I woke up one day and found that I was about to turn 30. I thought about the six years of hard work in the Internet industry. From Tencent, JD.com to ZhongAn Insurance. From category operation , event operation , user operation to data. From e-commerce to finance, I always want to learn from my predecessors and gain operational experience from actual work.

What is user operation ? This point is too big, it involves attracting new customers, promoting activation, and retaining customers. How to do user operations? Many people will tell you various ways and methods of Baidu search user operation, which seem to make sense, but are actually not very useful when you really apply them to your own work. Because that is a truth that the general public understands, and if everyone can understand it, the content is bound to be simple, and simple content will inevitably not contain much valuable information . This article focuses on the user lifecycle management that I do at Tencent, which is a very important aspect of user operations. If you can master the user life cycle model and effectively improve various data at work, congratulations, you are already a veteran in operations!

Introduction to User Lifecycle Model

We build the user life cycle model in two dimensions: horizontal and vertical.

Horizontally, we divide user maturity into primary users, growing users, and mature users based on user purchase frequency (the number of times a user shops). Vertically, we divide user warning status into active users, warning users, churned users, and dead users based on the time interval since the last purchase.

Frequency classification standard: the user retention rate increases greatly from the first purchase to the second purchase, but the retention rate does not increase significantly after the 8th purchase or more; last purchase classification standard: 80% of users will make a second purchase within 3 weeks, and 90% of users will make a second purchase within 5 weeks. (The above conclusions are based on platform data and cannot be applied rigidly)

How to determine the threshold for segmentation

User maturity: mainly based on the retention of each frequency to judge user maturity: It is found that users who purchase 1-2 times can basically be considered new users. Between 3-7 times, the retention rate will increase significantly with each additional purchase. The retention rate will decrease after 8 times and above, which means that users have basically become dependent on the platform.

User warning status: mainly judged based on the distribution of users with different last purchase times. 90% of users make purchases within 5 weeks. Users who make purchases over 5 weeks are basically considered to be lost. Users who do not make purchases over 6 months are basically considered dead. Users who have made purchases in the last three weeks have generated 85% of orders in the last 90 days. These users can be considered active users.

The basis for dividing active, declining and lost users: the last two order times of all users are counted, 90% of the users are within 35 days, and 80% of the users are within 21 days. So these two thresholds are defined as churn and decay respectively.

The basis for dividing primary, growth, and mature users: By counting the different cumulative purchase times of all users and comparing the weekly retention rates, it is found that the retention rate of users with less than 2 cumulative purchases is significantly lower, while the retention rate of users with more than 8 cumulative purchases tends to be stable, so these two thresholds are defined as primary and mature respectively.

User Distribution

After the user segmentation threshold is determined, we can see the distribution of platform users

User life cycle conversion direction

After the user life cycle model is established, it is important to use the model to improve the user conversion direction:

1. Recall lost primary users to active users;
2. Convert warning users among primary users into active users;
3. Convert primary users into growth users;
4. Transform growing users into mature users;
5. Recall lost users among mature users as active users;
6. Convert mature users who are alerted to active users;

Targeted operational plans

1. After placing an order, a coupon will be given in the floating layer. If you have not placed an order for 35 days, you will be recalled by SMS, and a large number of coupons will be sent, and a reminder that it has not expired (the last reminder will be sent by SMS)
2. After placing an order, a coupon will be given in the floating layer. If no order is placed within 21 days, PUSH will recall it, issue a general coupon, and remind you twice that it has not expired;
3. After a new user places an order, the coupon will be returned immediately, with a pop-up reminder, a red envelope with an expiration date, and a reminder twice that it has not expired; regular PUSH operation 2 to 3;
4. If no order is placed in 7 days/15 days, push messages based on consumption habit data to encourage more red envelopes;
5. When no order has been placed for 35 days, SMS recall will be sent, relevant new information will be promoted based on user behavior, combined coupons (categorized/limited-time coupons) will be issued, and reminders will be given twice if they have not expired (the last reminder will be sent via SMS);
6. If no order is placed in 21 days, PUSH recalls, sends new information based on the behavior, sends medium-strength coupons, and reminds twice that it has not expired;

Note: If the app is opened within 2 days before the PUSH/SMS, the floating layer can be used directly to replace the above PUSH/SMS contact channels .

Mobile application product promotion service: APP promotion service Qinggua Media advertising

This article was compiled and published by @旭说 (Qinggua Media) when it is reprinted. Please indicate the author information and source!