From the user's perspective, each product actually has a complete user life cycle (as shown below). The user system we often talk about is based on the user life cycle, modeling users, and then implementing many refined product solutions and operational strategies.

This article will look at how to systematically and systematically retain product users from the perspective of the user life cycle.

As shown in the figure, the yellow line roughly reflects the user value at different life stages within the product. We roughly divide users into the following categories based on their status in the product:

• New users : refers to those who have just entered the product and are about to or are in the process of experiencing the product functions.
• Core users : refers to the most stable group of users in the product, and also the most core users that our product needs to maintain
• Quasi-churn users : refers to core users in the product who show some signs of churn
• Lost users : Users who have been determined to be lost under the product business rules.

It can be clearly seen from the yellow line that the value of users in different states is different. The purpose of our retention is to convert more low-value users into high-value users .

We abstract the conversion path of user status and get the following picture:

We abstract the conversion path of user status and find that there are multiple conversion routes for the user's status in the product. The key to retention for us is to improve the effects of the three conversion paths 1, 2, and 3 in the figure.

The three paths correspond to users in three different states. Naturally, the retention strategies used are also different. Let’s take a look at them one by one.

Let's first list the users involved in the three paths:

There are three types of users: new users , existing users (core users + quasi-churned users), and churned users .

So, let’s take a look at the differences in retention for these three types of users:

1. Retention of new users

That is to say, after we attract users to the product through a series of means, how do we retain them? The core indicators are what we often call next-day retention, seven-day retention, monthly retention, etc.

2. Existing user retention

Our core goal is to increase the time users spend using our products or staying in our products through a series of measures, and prevent user churn. The core indicators are what we often call DAU, MAU, churn rate, etc.

3. Retain lost users

Our core goal is to use a series of measures to find ways to get users who have experienced our products and services to come back again. The most common core indicators are recall rate, or more in-depth recall user retention rate, etc.

So let’s take a look at these three scenarios separately.

New user retention

For the retention of new users, we usually look at one question:

Find the key nodes that affect user retention (also known as aha moments), and then guide users to complete "key behaviors" to make them perceive the value of the product, thereby improving the retention of new users.

What is the key user behavior (aha moment)?

The Aha moment is when a user first realizes that the product is valuable to them.

There is a standard formula for user key behaviors (aha moments):

Who completes what behavior , how many times, and at what time?

Finding the Aha moment actually means finding the behavioral differences between active users and churned users through analysis, and analyzing the core demands of users behind the behavioral differences; then, through product or operational means, the core demands of new users are met as much as possible, thereby retaining new users.

Then we need to clarify how to find the aha moment? That is, the key behavioral nodes that affect retention.

Step 1: Extract product value, sort out possible key user behaviors, and find the fastest way for new users to feel long-term value when they start using the product

For example, we can use a few questions to find the value of the product:

• who: who is the user
• What: What problem do users want to solve with this product?
• why: Why should the user solve this problem?
• Vs: What other ways do users have to solve this problem?

Then, through user surveys, we compare the responses of different users to find out the most important value of the product to users and find a series of key behaviors.

• For the most active long-term users: Why do you find the product valuable?
• For users who leave quickly after registering: Why do they leave quickly?
• For users who actively use the app after registration: Why do they stay? What actions did you take as a new user and what were the key experiences?

Step 2: Through data analysis, find the key behaviors that have the strongest positive correlation with user retention among the above key behaviors

Translated into plain language, it means how to find the most valuable behavior from a bunch of seemingly valuable behaviors.

One of the methods we often use here is to compare the retention curves. The same comparative experiment thinking is used to control a single variable to compare the user retention data of users who did and did not perform the behavior. Among them, the behaviors corresponding to the two sets of data with the largest difference are most likely the key behaviors we are looking for.

Remember the formula for the aha moment we talked about above?

aha moment = who does what , how many times, and at what time

The “ who ” and “ behavior ” have been found in the first and second steps. Next, let’s look at the “ how long ” question.

Step 3: Find the effective cycle of key user behaviors

Guiding users to complete key behaviors is time-sensitive, and talking about results without considering the time is irresponsible.

For example, on a certain social platform, the key behavior for user retention is to post once. However, if a user only posts once a year after registration, can we say that the user has been retained by us? It doesn't seem reasonable, but more like a recall.

Of course, the issue of timeliness needs to be combined with the specific business logic of the product. Assuming that your business logic is indeed that users only need to come once a year, it is reasonable for users to complete key behaviors within one year.

For example, a product that everyone may be using is the "Personal Income Tax App", but perhaps everyone only uses it when they do their annual tax reconciliation at the beginning of the year. The “how long” here can be within a year.

So how do we find out “how long” for our own products? There are several principles for reference:

Principle 1: The higher the frequency of use, the faster the activation needs to be

The higher the frequency of use, the sooner new users expect to get value from the product. You can roughly determine the activation period of new users based on the frequency of use.

Here are some examples:

• For social, short video, and gaming products, users use them about every day. If you fail to let users discover value on the first day, there is a high probability that users will be lost. Therefore, the “how long” for social, short video, and gaming products is ≈ 1-3 days
• For products like food delivery and fitness, users use them about once a week. So at this time, we have to find a way to let users experience the value of our products as much as possible within a week. How long? ≈ 3-7 days
• For e-commerce products, users use them about once a month, so there is more room for maneuver. How long? ≈ 7-30 days

Principle 2: The shorter the life cycle, the faster the activation needs to be

• The shorter the product life cycle, the sooner new users expect to get value from the product
• For example, the product life cycle of games is relatively short, so the activation period is relatively shorter.
• The life cycle of e-commerce products is relatively long, so the activation period is relatively longer.

Principle 3: Refer to actual data

Analyze the actual data of new users to see the time window when most early activation behaviors occur

• For example, for a community product, we pulled out the time distribution of all users who published content for the first time and found that 80% of them happened within a week, so we can define "how long" as ≈ 7 days

Then come back to this formula

aha moment = who does what , how many times, and at what time

We’ll talk about “ how many times ” later, let’s first look at “ behavior ”

Step 4: Find the optimal number of times for this key behavior through data analysis

Why calculate the optimal number of reps?

Some activation behaviors only need to be done once, such as the first order of e-commerce; some activation behaviors need to be repeated many times to ensure that new users feel the value of the product, such as watching a video.

Theoretically, the more repetitions you do, the greater the improvement in retention. However, the activation time for new users is limited, and it is unrealistic to ask users to repeat too many times. Therefore, we find the optimal number of activation behaviors to ensure that users gain value without burdening them.

How to calculate?

Find the value that has the largest marginal effect on the number of behaviors on the retained data. Let me explain this with an example:

Suppose that the key behavior that affects retention of a product is posting, and the behavior cycle is one week.

Then we first draw a distribution chart of the number of activation behaviors of new users on the first day (the number of users with different posting times within a week)

Then, analyze the relationship between the number of key behaviors in the first week and the retention rate in the second week (as shown below):

Finally, we find the point with the largest retention editing effect (usually the inflection point of the curve). The number of key behaviors corresponding to this point is the optimal number we are looking for (as shown below)

OK, so far, we have a pretty clear idea of ​​what the aha moment is.

aha moment = who does what , how many times, and at what time

However, a special reminder is needed here. Through the above series of operations, we actually found the correlation factors that affect retention, rather than the causal factors.

What is correlation?

It is observed that users who have a certain behavior have a higher retention rate

What is causality?

The user performed a certain action, resulting in a higher retention rate

Correlation can only help us analyze and experiment, but causality can help us know when to perform actions .

Therefore, after we obtain a series of correlation conclusions, we still need to continue to optimize and experiment to ultimately find the causal factors.

User behavior analysis

Once we find the aha moment that affects user retention, how do we guide users to complete this key behavior?

By analyzing user data, we can find the key path for new users to complete their behaviors and analyze the path conversion funnel.

Let’s take a community product as an example. Suppose the path for users of this product to complete key behaviors is as follows:

[ Open the app->browse the content in the square->find the topic you are interested in->open the theme square->browse the content in the theme square->publish your own content->publish successfully ]

Then for this path, we can do analysis in several dimensions:

1) Path analysis

What is the actual path for new users? Is it interfered by irrelevant paths?

For example, can users post directly in the square without entering the corresponding topic, and then just choose the topic when writing the content?

According to this idea, we can optimize the user path:

[Open the app->browse the content in the plaza->publish the content->select the topic->publish successfully]

For users, there are fewer steps to go through, so the conversion rate of the overall path will naturally increase.

2) Detailed track inspection

What actions do the lost new users take in the product?

We can look specifically at which link in the path has the most serious loss, then analyze the reasons for the loss and solve them in a targeted manner.

For example, the user churn rate is relatively high after the action of "browsing the square content", because our next step is to guide users to find topics of interest. The reason for the high churn in this step may be that users did not find the topics or content they were interested in.

To address this issue, can we guide users to select a few tags of interest when they first register, and then prioritize pushing content that is more in line with their interests after they enter the product?

3) Behavior funnel time interval

How big is the time interval window between user path nodes? How long is the time interval between each step?

For example: [ Browse the content in the theme square -> Publish your own content ], the time interval between this link is relatively long. So can we provide some guidance on publishing this action?

For example, when users are browsing content in the theme square, can the reminder "Interested in this topic? Let's talk about your opinion!" accompanied by dynamic effects guide user behavior? This may be one of the methods. When users are not motivated in a certain behavior, we need to find a way to stimulate their motivation.

4) User Segmentation

Is the user churn rate the same for different segments? If they are different, they need to be grouped and corresponding strategies implemented.

This is actually a combination strategy of the above three dimensions based on different user attributes. We can group users with the same attributes, and then conduct analysis and strategies in the above three dimensions separately. I won’t say much here.

Finally, let’s summarize the user behavior guidance:

1. Understand the user's real path through path analysis, build a funnel, and analyze the conversion rate of each link
2. Through user segmentation, can we understand whether the churn rates of different segments are different? (Common segmentation dimensions: user portrait, customer acquisition channel, device platform, product line, CRM channel, red envelope subsidy, demography, customer service interaction, community interaction, etc.)
3. Understand the user's behavior progress through behavioral funnel interval analysis
4. Understand the reasons behind this through user research
5. Understand what elements new users click first with click heatmaps

Through the above steps, we can basically analyze the reasons for user loss on the user behavior path. The next step is to optimize the product in a targeted manner.

How to optimize? Continue reading below~

User behavior optimization

In the product, there is such a formula for studying user behavior:

Behavior = (motivation - resistance) * nudge + reward

• Behavior: The behavior you want users to complete
• Motivation: How much the user wants to complete the action
• Resistance: How easy is the behavior to do?
• Nudge: Prompt users to take action
• Reward: What kind of feedback can users get after completing the behavior?

Our direction for optimizing user behavior actually revolves around these directions. Then let’s take a look at them separately

Behavior:

It is the behavior that you want users to complete, which is the "key user behavior" we mentioned above. How to find and analyze it has been explained thoroughly above, so I will not elaborate here.

power:

Users want to complete this behavior very much, so what are the factors that affect user motivation?

• How strong and urgent is the user's need to use the product?
• Do users have any other alternatives?
• How long does it take for users to make a decision?
• Are the user and the decision maker the same person?

So how do we improve users’ “motivation”?

Explore the user's own needs and provide assistance to make the user more willing to complete the behavior

Introducing several common methods

1) Leverage social connections

Social connections are the best way to build trust and personalize the experience for new users

Common cases:

• WeChat Reading: After a new user logs in to WeChat, the homepage shows the books that their friends are reading
• Uber: When a new user registers, if the user is recommended by a friend, the friend’s name will be displayed.

2) New user bonus

Red envelopes and incentives should be related to the core business goals of the product, and efforts should be rewarded. At the same time, users should be prevented from taking advantage of the product.

Common cases:

• NetEase Yanxuan New Member Red Packet: New Member Coupon: Increase ordering, and receive the mobile phone number required
• Daily Youxian New Customer Discount 100 Yuan Coupon: Increase Order Rate

3) Explain why

When making a request to a user, tell her why and what benefits it will bring to her.

Common cases:

• Maimai: Allowing users to share address book permissions means that it can improve career growth opportunities

4) Personalized product experience

Personalization gives users what they want, and is also a way to increase motivation so that users can choose their own preferences.

Common cases:

• Xiaohongshu: The new user registration process allows users to select the content they are interested in

5) Simulate the pre-Aha moment

If it takes a long time for a new user to experience the Aha moment, you can simulate the Aha moment and put it in front of the user.

Common cases:

• Zhuanzhuan: When recycling mobile phones, users can get an online valuation first, so that they can see the benefits and increase their motivation.

6) Influence users through user psychology

Increase users' motivation to act quickly by creating a sense of tension, scarcity, and gamification

Common cases:

• WeChat Reading: How many new user unlimited cards are left today?
• Online education: Course pricing is tiered, with a 50 yuan increase when 50 or more people sign up.
• Pinduoduo: Remaining time of coupon validity

resistance

How much do users want to complete this behavior? What are the possible consequences for users?

1) How difficult is it for users to feel the value of the product ?

• Are the functions of the product easy to understand and use?
• Is it a long or short time for users to reach the Aha moment?
• What are the capabilities and qualities of the users themselves?

2) Physical barriers in the new user flow

• Do I have to register to use it?
• Are there many steps in the activation process? Is there any interference?
• Does it require a lot of information? Is sensitive information required?
• Do I need to pay to use it?

3) Cognitive barriers in the new user behavior process

• Is the onboarding process and copywriting confusing?
• Are new users given too many choices or a cold start?
• Is the information inconsistent, confusing users?

How to reduce resistance?

Remove all obstacles that prevent users from completing activation actions and help users quickly reach the Aha moment

Introducing several common methods

1) Remove unnecessary steps and information

In the conversion process, one more step means one more loss. Optimizing the process can reduce the loss

Common cases:

• User login and registration: WeChat one-click login, one-click registration

2) Avoid user cold starts

Give users default options and content, don’t leave it blank at the beginning

Common cases:

• Investment APP: Add 10 yuan default value

3) Highlight key behaviors and paths

Prioritize users to use core functions and highlight key functions

Common cases:

• Beginner's guide for logging into the app for the first time

4) Avoid giving users too many choices

Too many choices may cause users to not choose any

Common cases:

• Investment apps: Only give new users 1-5 investment options

5) Use first and then pay or register

First, let users experience it at the lowest cost, enhance user trust, and then guide subsequent actions

Common cases:

• Free use for the first month, 2 free courses for new users

Boost

If users want to complete this behavior, what kind of help can we provide them?

Some ideas you can refer to are:

• See how fast users make decisions?
• How long is the time window to win back users?
• If the user is not activated, is there any channel to reach the user?

How to boost?

Use multiple means to help users complete activation within the critical time window

Introducing several common methods

1) Newbie Guide

It is important to provide new users with guidance at the right time through various means, and do not let new users' guidance become an obstacle for users to use the product.

Common cases:

• The novice guide boot page of the curtain
• Key function pop-up window guidance after entering the product
• Product documentation

2) User recall

If the user does not complete the activation action, use external channels to promptly pull the user back to the product through push notifications and other means to continue trying the product until the Aha moment.

Common cases:

• E-commerce platform non-payment reminder

award

If helping users complete a behavior is the push, then giving users the reward for completing the behavior is the pull.

When designing rewards, we need to consider:

• The more difficult the action is for the user to complete, the more reward they should receive after completing it.
• If the behavior is a key action for your product, rewards can help form a habit
• If the process is too long, give users rewards in the middle to recharge

How to reward?

Provide timely feedback and rewards to users who have completed activation behaviors, encouraging them to move forward and complete more behaviors.

Introducing several common methods

1) Timely incentives

After the user completes the key behavior, celebrate to show encouragement. The more information required, the more difficult the behavior, the greater the encouragement should be, and reward the user with surprises.

Common cases:

• Maimai: After completing registration, you can mine the network information animation

2. Timely feedback

Give users instant feedback After the user completes the action, timely feedback, even if it is small, makes the user feel rewarded

Common cases:

• Xiaohongshu: Generate homepage animation after completing user evaluation
• 58 Tongcheng: Use progress bar to show progress

To sum up, there are three main steps to retain new users:

Identify the key behaviors that affect new user retention -> Sort out the paths for users to complete key behaviors -> Improve the conversion rate of each node on the behavior path

That’s all about “retention of new users”. Next, let’s study “Retention of Existing Users”

Retention of existing users

As mentioned earlier, our core goal for retaining existing users is to increase the time users spend using our products or staying in our products through a series of measures to prevent user churn. The core indicators are what we often call DAU, MAU, churn rate, etc.

There are two commonly used methods:

1. Continuous stimulation

Continue to explore other needs of existing users and design new services or functions to meet the diverse needs of users

For example:

• Meituan: The initial core function was to provide group purchase services for offline consumption scenarios. Later, a series of other business scenarios such as taxi-hailing and food delivery were added to meet the diverse needs of users and allow users to continue to experience different values ​​in the product.
• Alipay: Its initial core function was to serve as a transaction guarantor between merchants and buyers. Subsequently, it launched business scenarios such as financial management and life services, which also allowed users to continue to feel different values ​​in the product.

Well, around this method, in fact, our general idea of ​​product design is the same:

Find users -> analyze user needs -> find new pain points -> design functions to satisfy them

2. Develop a habit

Guide users to complete core behaviors multiple times to strengthen user minds. Let's focus on this one. Compared with the first method, this one is a scenario we encounter more often in our daily lives.

To help users develop a habit, we have to mention the word that has been "overused": HOOK model

The hook model was proposed by the author of "Addicted" and is an introduction to how to make users addicted to a product and form habits.

As shown in the figure, the HOOK model can be roughly divided into four processes: trigger + action + reward + investment

The essence of the HOOK model is to guide user behavior and allow users to complete a behavioral loop in the product content. Because this model is quite versatile, we will use it here to discuss specifically how we can retain existing users.

trigger

Triggers are divided into external triggers and internal triggers, which guide users to take actions based on different user scenarios.

• External triggers generally refer to external operational actions or marketing activities. Such as online activities, offline advertising, public account posts, etc.
• Internal triggers generally refer to amplifying users’ deep-seated needs by guiding their emotions. For example, negative emotions: "I feel like I'm falling behind" positive emotions: "I like the feeling of progress"

To sum up, external triggers mean that users have needs, and we let users see them through channels that they may see. Internal triggers mean that the user’s own needs may not be very strong, so we use some emotional and psychological guidance to amplify the user’s emotions.

action

We need to know that for products on any platform, if you want users to take action, you can’t do without a formula:

B (action) = M (motivation) + A (ability) + T (trigger)

Let’s look at these three factors separately.

The first is motivation . There are several common motivations:

• User "feelings": For example, users want to pursue happiness, or users want to avoid pain
• User "expectations": For example, users want to gain hope, or users want to avoid disappointment
• The user's "sense of belonging": for example, the user wants to be recognized by others, or the user wants to avoid being rejected in person

How to improve user motivation? We generally do this from two perspectives:

• Improve user internal motivation

The scenario is usually that the user has certain "emotions" and then we amplify them.

For example, there is a promotional slogan that is widely circulated on the Internet: "If you come, we will train your children; if you don't come, we will train your children's opponents." It may contain some ridicule and jokes, but it is a very typical means of amplifying user anxiety.

In addition, there are some content platforms that have introduced various creation incentive strategies, which actually fall into the category of enhancing user motivation.

• Create external motivation for users

This scenario is usually that the user has no motivation yet, so we try to create motivation for the user.

I don’t know if you have seen some Python course advertisements recently. The angles are so novel that they are quite shocking.

Originally, this is a programming language, but in advertisements it seems to have become a necessary skill for people in the workplace. Product learning can help people do better data analysis, operations learning can help people better monitor activities, HR learning can help people better collect statistics, and those who create content can also crawl a lot of information. This is a typical case of having no motivation to create.

There are also other products, such as Qutoutiao, Douyin Express, Weishi, etc., which have strategies such as getting cash for reading and getting cash for watching videos. These also fall under the category of creative motivation.

Then comes the ability

Improving users' ability to act can also be analyzed from two perspectives. They are to enhance user capabilities and lower the threshold for action.

Improving user capabilities, often targeting "small B users" in products, such as: short video creation guidance for Douyin, store opening training for merchants in e-commerce products, etc.

It is common to lower the user threshold, such as

• Focusing on reducing user operation time: Bilibili’s one-click three-link function, password-free payment for various transaction products, fast payment, etc.
• Focusing on reducing user's operating effort: one-click typesetting function of content platforms, and the "re-order" function of food delivery platforms
• Focusing on reducing users’ mental workload: e-commerce platforms display discounted prices and optimal combinations, and food delivery platforms display discount combinations
• Focusing on reducing users’ expenditure costs: instant discount on first order, installment payment, etc.

Finally, the trigger

The trigger here is basically similar to the trigger in the hook model discussed above, so I won’t go into details.

(It should be noted here that the user behavior motivation formula and hook are two parallel methods, which are used interchangeably here mainly to facilitate the explanation of the point of view)

award

Let's take a look at rewards. Here we also divide the types of rewards into categories: rule rewards, social rewards, and spiritual rewards.

• Rule rewards: mainly refers to giving users clear behavioral feedback, such as common animation effects and sound effects
• Social rewards: Use social relationships to get other users to give feedback, such as likes, rewards, and attention.
• Spiritual rewards: meet the spiritual needs of users, such as learning, ranking, identity, charity, etc.

Investment

Let users invest time, energy, money, data, and emotions, and become more and more dependent on products

Common investment methods:

• Time: Cultivation functions, such as Ant Forest
• Money: krypton gold products, such as games, membership
• Energy: Output products, such as articles and news
• Data: Recording products, such as fitness, study, and schedule
• Emotion: social products, fans, follow lists, and friend relationships

The above is a brief introduction to the hook model. There is one more point that needs special explanation:

Before using the hook model, we need to analyze user behavior. The simplest thing is that we must first find effective habitual actions. This is the same logic as finding the "key behavior" mentioned earlier, so I won't go into details.

Regarding the retention of existing users, we also summarize

To retain existing users, we can take two approaches: continuous stimulation and habit formation.

If retaining new users means we need to find the most suitable user path, then retaining existing users means we need to find the optimal behavioral closed loop .

Based on the common behaviors of habitual users, we can sort out the paths that users take to complete these actions and find all the key nodes on the paths.

Then check whether there is a corresponding behavior closed loop. If so, find the optimization space for each node. If not, add nodes to form a closed loop.

Finally, through the closed loop of behavior, users are repeatedly stimulated, allowing them to continuously experience the path and eventually form a habit.

Lost user retention

Finally, let’s take a look at the retention of lost users. The process of retaining lost users can be roughly divided into several steps:

Define lost users –> Develop an early warning mechanism –> Intervene in impending churn –> Recall lost users

Step 1: Define Churning Users

We need to have a clear standard to determine what kind of users are lost users. For example, the definition we often use is: if a user has not performed any action for a certain period of time, it is considered a lost user.

The specific "how long" and "what behavior" here need to be determined based on the specific product business scenario.

For example, for an e-commerce product, we define a user as a lost user if he or she has not made any purchases within one month.

Step 2: Develop a loss warning mechanism

We can understand the early warning mechanism as a warning line, which means that the user is in a state of "about to be lost". Why do we need an early warning mechanism?

Regardless of whether they are lost users or users who are about to lose users, our ultimate goal is to guide users to become core users again. As for becoming core users, it is obvious that the conversion path of users who are about to lose is shorter than that of users who have already lost, and conversion is relatively easier.

As for churn warning, it is also a process of judging user behavior. If a user exhibits certain characteristics or behaviors, there is a high probability that he or she will be lost.

Step 3: Intervene with users who are about to churn

If there is a chance to rescue, try to rescue, and avoid loss if possible.

At this time, we need to quickly determine the reasons why users are about to leave and then solve them in a targeted manner.

There are many categories of reasons for analysis, if they are just some relatively easy to solve, or they are problems with user operations. We can improve in time. At this time, it is necessary to follow up in a timely manner and synchronize the problem improvement status to the user. This will most likely help recover lost users.

What if it is a bigger problem, or a problem that cannot be adjusted in a short period of time? At this time, we should also keep records of which users have this problem. After we correct these users, we can do some targeted recall actions.

Step 4: Recall lost users

So for the users who have already lost, let’s see how to recall them.

One of the methods was mentioned just now. It is to analyze the reasons for user loss. If the problems encountered by the users have been solved, then targeted recall can be carried out.

If we have not yet determined the reason for user churn very accurately, then at this time we can take some recall measures based on the churned user data we already have.

For example, analyzing the user’s past core behaviors and analyzing the user’s current possible demands. Then develop a strategy.

For example:

• For example, if the core behavior of a user in our product is to frequently post articles, then we can push the likes and reading status of the articles;
• For example, if a user frequently watches videos in our product before, we can push the latest works of the bloggers he follows to the user.
• For example, when a user churned a year ago, we had a label that he had a 5-grade child, so a year later we could push some high school entrance examination courses.

This is the formulation of a recall strategy, and the core is to look at the current needs of users. Another thing that needs to be confirmed is the contact method and materials used when recalling users.

• The contact methods are relatively common, such as SMS, email, push, official account, etc.
• There are also several categories of materials that can be used for outreach : text notifications, welfare distribution, social relationships, etc. You can choose the most appropriate one based on the product category.

Ok, this is my basic understanding of recalling lost users.

Summarize

Finally, let’s summarize it as a whole. Regarding user retention, I classify users from the perspective of the user life cycle, and then formulate strategies based on the different demands of users in different states.

As shown in the figure, our core is to optimize and formulate strategies for three of the paths.

Path 1: Improve the conversion rate from “new users” to “core users”

After we attract users to the product through a series of means, how do we retain them?

Our common approach is to find the key behaviors (aha moments) for retaining new users, and then optimize each user path focusing on the key behaviors.

Path 2: Turn users who are about to churn back into core users

For the existing core user pool, how can we reduce user churn?

We can consider this issue from two perspectives. The first is to increase the points where the product meets the needs of users, so that users are continuously "stimulated" and increase the frequency of users using the product. The second is to cultivate users' habits of using the product, increase the time users spend on the product, and make users "addicted".

Path 3: Recall lost users

At this time, we need to use a series of measures to find ways to make users who have experienced our products and services come back again and feel the new value again.

There are also two common analysis angles: the first is to analyze the reasons for user churn and solve user problems in a targeted manner; the second is to analyze the user’s current core demands and provide solutions.

Author: kiwi

Source: Hutong No. 6