How to build a user operation system?

As the product size grows, the number and types of users become increasingly complex and diverse. At this time, a single operating strategy is no longer sufficient to support product operations. The author provides us with a new idea and a new operation strategy to rebuild the user operation system and effectively solve this problem.

1. Background

During the early operation of the product, because the number of users was small and user behavior was unstable, we designed many scattered operation strategies in the product to guide and activate users according to different business purposes.

As the product continues to develop, its functions increase, the number of users grows, and the types of users become more diverse; however, the product's operating strategy becomes more complex and bulky, which is not very friendly to subsequent new users and is not conducive to the organization of operating ideas and the formulation of operating strategies.

Therefore, it is necessary to sort out the scattered operation strategies and, with the support of the existing user data accumulation, rebuild a complete, clear and effective user operation system.

2. Build the framework

3. Detailed steps

1. Sort out product commercial channels & user value growth paths

1.1 Commercial Channel Analysis

• DAU value: We expect large-scale user stickiness and continuous activity, and a user habit formation channel is needed within the product.
• GMV value: We expect users to have good payment habits, and there needs to be a payment behavior conversion channel within the product.

1.2 Analysis of User Value Growth Path

2. Segment users based on user value & personality roles

2.1 Stratification based on user roles

By breaking down and analyzing the product, we can qualitatively assign roles to different types of users in the product based on experience.

2.2 Layering based on life cycle

The life cycle is the entire process from the user [contacting the product] to [leaving the product]. The first task is how to divide the life cycle?

Through the above business channel analysis, we know that the commercial value of the product is reflected in DAU and GMV, so the user life cycle also needs to be divided according to these two values:

Extract the retained users who have registered for more than 15 days, capture their login times + usage time + payment amount within 15 days for analysis, and find out the high-value users.

Group the retained users who registered 15 days ago:

Observe the distribution of login times, average usage time, and spending amount of this group of users.

Through data analysis, we found that the data of this group of users who have been retained for more than 15 days are as follows:

• 31% of users log in 1-5 times
• 49% of users log in 5-8 times
• 20% of users log in more than 8 times

• 11% of users have an average visit time of less than 30 minutes
• 52% of users have an average visit time of 30-120 minutes
• 37% of users have an average visit time of more than 120 minutes

• 57% of users paid less than 100 yuan
• 29% of users paid between 100-500
• 14% of users paid more than 500 yuan

Based on the common characteristics of retained user behaviors analyzed above, users are divided into life cycles:

Combine the previously defined personality role stratification with the life cycle stratification, divide the users in the growth stage and the mature stage a second time, break down the granularity of the operation further, and build a growth ladder for users to facilitate more targeted and refined operations later.

3. Build a user growth ladder + operation model based on user stratification

3.1 Build a user growth ladder

By combining the previously defined personality role stratification with the life cycle stratification, the user's value is increased layer by layer.

3.2 Sorting out the user operation model based on the user growth ladder

4. Sort out the growth paths at different stages of the growth ladder and screen them (user retention and funnel analysis)

Through the Sangji diagram, all growth paths of users at different stages can be enumerated, and the paths that can achieve the goal of improving the user's growth stage can be listed:

At each growth stage, several improvement paths are sorted out, and through funnels, retention analysis, etc., the growth paths with operational value at each stage are screened out (multiple paths are allowed) .

Taking the process from registered users to novice users as an example, we analyze how to screen out valuable growth paths. The user group analysis process at other stages is similar.

We first analyzed the page access path and combined it with the business to discover two paths for users to advance from registered users to novice users and then to use the core functions of the product.

Next, we analyze these two paths. Considering that both paths are main paths, in order to make the user's behavior in the early stage meet the expectations of product design as much as possible, so as to avoid too many choices and lead to loss, we only keep one of the two paths; and focus on doing a good job in related operational polishing and guidance work. The analysis mainly focuses on three aspects:

1. Which path is better?
2. What characteristics do most users meet?
3. Do most users perform the same actions?

Sort out all the growth paths from new users to novice users:

Analyze relevance (which path is better → analyze basic commonalities → analyze behavioral commonalities)

First, use funnel analysis to compare the conversion rates of the two conversion paths.

Path A Funnel:

Path B Funnel:

Next, the conversion population and loss population of these two paths will be saved separately and compared and analyzed, mainly using the user's basic data + user behavior data as the analysis dimensions to see if some commonalities can be discovered to provide decision-making for subsequent path optimization.

Next, we use the event analysis function based on the user attributes to analyze these users in multiple dimensions, mainly from the aspects of region, gender, age, etc.

Next, we analyze the user's behavioral characteristics and use the event analysis function to conduct retention-related analysis on the group of users who have successfully converted, in order to test the subsequent stability of the conversion, and to discover the key conversion functions for retention and key functions for improving value.

Evaluation results (comparison of conversion rate, comparison of retention rate):

1. Among the 2,000 new users, 1,238 users completed Path 1, with a conversion rate of 74%; 1,738 users completed Path 2, with a conversion rate of 85%. So the 2 paths are better.

2. Among the more than 3,000 users who successfully converted through the two paths:

• a. The male to female ratio is 77.3% and 22.7%.
• b. Users in first- and second-tier cities account for 32.3%, and users in third- and fourth-tier cities account for 67.7%.
• c. Among users who used the "Follow the anchor" function, the second-time retention rate was 30%, and the average on the site was 20%.

After analyzing all the paths of different improvement stages in the user growth path table, you can gain some understanding of these paths and the results of data analysis. These analysis results can be organized to provide decision support for subsequent operational strategies.

5. Implement operational strategies based on user growth paths

Through the above series of analyses, we have gained insights into improving users at different levels. Next, we need to use these insights to design operational strategies for different nodes of each path. That is, to guide and motivate user behavior, let users follow our preset path, and thus improve user value.

Before designing an operational strategy, user behavior needs to be classified into:

• One-time behavior (generally exists in the main path, that is, the first time a certain goal is achieved, such as the first gift giving, the first broadcast, etc.).
• Milestone behaviors (mainly users incentivize users’ cumulative behaviors, such as signing in for 10 days, getting more than 1,000 likes, being online for 30 minutes, etc.).
• Continuous behavior (actions that require users to perform continuously, daily login rewards, each broadcast, etc.).

Find the incentive points in the user's growth path that require one-time incentives or milestone incentives.

Find the incentive points that require long-term and continuous incentives from the product function modules.

Operation strategy design:

6. Design a level system to connect the user's growth path

Assign rewards and growth values ​​to motivated user behaviors, build user level curves according to different function models, and finally match the corresponding user levels to rights and rewards.

Growth system design: omitted

7. Design of pre-churn warning and recall mechanism

7.1 Define churn indicators and identify churn groups

By analyzing the user's return rate, we found that when the return rate drops to 5%, there will be an obvious turning point, and the data will tend to be flat afterwards. Therefore, users who have not logged in for 30 days are defined as lost users.

7.2 Construct a structure diagram of lost users at different levels and analyze lost user behavior

• Is it a problem in the process? Analyze the proportion of lost users based on the novice period, growth period, and mature period.
• Is it caused by user quality issues: Analyze the proportion of lost users from different channels.
• Whether it is a specific user group issue: Analyze the basic attributes of users, such as gender, region, age, interests, etc. to see if they are similar.
• Is it a specific behavior/function problem: Analyze the churn scenario and what behaviors the user performed/what functions the user used.

Below, we take the novice period as an example to analyze the behavior of lost users. The analysis process of user groups in other stages is similar.

Analyze the characteristics of lost users and establish a graded early warning database for lost users.

Grouping users:

Analysis of the average daily usage time of novice users:

Analyze the number of logins per week:

The previous analysis observed that the "Follow the anchor" function can greatly improve retention, so this function is also analyzed to analyze the number of anchors that users follow.

Visualize the data results:

Based on data analysis, the characteristics of user loss during the novice period are summarized as follows:

• 55% of users come from QR codes promoted on the ground
• 98% of users use the app for less than 30 minutes a day.
• 93% of users log in less than 2 times a day
• 92% of users follow less than 2 anchors

7.3 Monitoring data and modeling the users who are expected to churn during the beginner period

• The natural churn rate of users registered through offline marketing is high and their quality is poor.
• Users with less than 2 daily logins for 3 consecutive days are marked as pre-churn users.
• Users who have passed the registration period and are active but do not follow the anchor are marked as pre-churn users.
• Users whose daily online time is less than 30 minutes for three consecutive days are marked as pre-churn users.

…

7.4 Formulate recall strategies and establish an automatic recall coordination system

Recall strategy: N/A

Related reading:

1. User operation: new funnel model for conversion analysis!

2. User operation: How to use B-side operation thinking to increase user growth?

3. Product operation: How to use data analysis to drive product user growth?

4. APP user growth: One model solves 90% of growth problems!

5.How to increase users? Take Pinduoduo and Xiaohongshu as examples

6. Triggering user growth: Is user operation just about attracting new users?

7. User operation: What else can you do to attract new users without fission users?

8. User operation: how can financial products awaken dormant users?

9. User Operation | How to perform user behavior path analysis?

Author: AFen

Source: AFen