When increasing user growth, how can we retain new users?

Retention of new users is a very important part of user growth , but it is easily overlooked. This article will talk about the importance of retaining new users and the execution framework ideas.

1. What is User Growth?

Let me first talk about what user growth is. My understanding may be different from others. From the perspective of foreign concepts, user growth is the AARRR model (as shown below).

From a domestic perspective, part of the value of user growth lies in the fact that it is a set of working methods and a form of team collaboration.

Moreover, because the domestic Internet environment and user habits are very different from those abroad, the understanding of the three aspects of acquisition, activation, and recommendation in the AARRR model is different at home and abroad.

For example, there are no operations or channel teams abroad. The main focus is on products, R&D, and marketing. There are not so many channels abroad, nor is there such a diverse user structure as in China. The difference is too big. This is why you see a lot of awesome cases abroad, but they are completely unfeasible in China.

Therefore, we can use the concept of user growth, but we must do a good job of "localization".

User growth first requires a set of working methods .

For example, if I am the boss of a company, I hope that the DAU of the app can be increased so that the company can make more money through commercialization. This is the overall goal.

The event operator told me that his goal for the event is UV. If the number of UV in the event increases, the DAU will increase.

The user operation staff told me that his goal is to maintain more core users and allow them to contribute more or be more active. With the drive of core users, DAU will increase.

The product manager told me that his goal is the UV of the product module he is responsible for, or its share in DAU. The higher these two values ​​are, the higher the DAU will be.

But in my opinion, the three figures of active UV, core user activity, and DAU percentage of a certain product module are completely not in the same direction and are not on the same dimension. It’s like everyone pushing a cart together, but if we don’t push in the same direction, the efficiency will be very low.

User growth ensures that the entire company, or at least a group of people, only do things that can clearly bring growth, and the direction and goals are very consistent. Don’t do anything that doesn’t drive growth or can’t be quantified.

User growth, and secondly a form of team collaboration.

The traditional Internet team divides the work as follows: product managers, operations, R&D, design, marketing, channels, and sales are each independent teams, each managed by a boss.

This kind of division of labor is simply "clustering", putting people with the same type of work together and then finding a more capable person to be responsible for management. Apart from this, there seems to be no other reason.

But in reality, people with the same job rarely collaborate with each other, and no matter how many product managers get together, they cannot launch a function. Therefore, the collaboration is actually more about product + operation + design + R&D, such a project team.

However, these roles in different teams may have different KPIs, which will bring about collaboration costs and reduce work efficiency.

On the other hand, such a vertical team setting will also bring about the "silo effect", as shown in the figure below. Each team works independently, and vertical collaboration may be deep enough, but there is little horizontal collaboration. If everyone only looks at his own little piece of land, it will also affect work efficiency.

User growth is about connecting departments together through a form of cross-team collaboration with the same goal to break the silo effect.

Image source: Hiking Artist

There are two common misunderstandings in understanding user growth.

First of all, user growth is not a secret or a shortcut. Bosses should not bring this concept into their own teams and then wait for data to improve rapidly. This is not a matter of one move or one technique, nor is it that mysterious. It is just a way of thinking.

Secondly, don’t be too narrow-minded and think that “fission” is all about growth , which is a generalization. User growth is the A AR RR model, and fission is one of the first A methods. In other words, there are many forms of acquiring customers. Previously there were channel delivery, event-based customer acquisition, etc., and now there is an additional form of fission.

There is a phenomenon of promoting the "fission" gameplay as "user growth" and making profits from it. Please do not be misled.

2. Why retain new users?

In the AARRR model, the retention of new users is the "activation" step (as shown above), but few people in China call it that.

Why is it important to retain new users? Let me tell you a story (picture above).

Imagine that you opened a small restaurant. When the first table of guests came, you would definitely do your best to treat them with the best service and the best food. When they leave, you might even give them a coupon for a discount next time they come.

Because you know, this first table of guests is very important. If they are satisfied with the meal, they will come back to consume multiple times, as customers accumulate over time. They will also recommend it to people around them, as restaurants are supposed to serve the people around them. They may also give you some feedback to help you improve your service.

This first table of guests is a new user for you. Their first experience is very important, and that's what you need to focus on.

Similarly, it is very important for us to retain new users of our Internet products. There are two reasons:

1. Good retention of new users will lead to an increase in overall retained users and avoid wasting customer acquisition costs.

For example (see the figure below), there are two channels to acquire users, and the number of new users brought in every day is 10,000. The next-day retention rate of channel 1 is 40%, so there are only 4K active users on the second day; the next-day retention rate of channel 2 is 30%, so there are only 3K active users on the second day.

Then the total number of active users brought by channel 1 is the yellow + black area in the figure below; the total number of active users brought by channel 2 is the black area in the figure below. It can be obviously seen that the former area is larger than the latter.

The decisive factor for the size of the area is the point of retention on the next day (the retained users on the second day in the figure, the green dot). Therefore, secondary retention is very important and determines the overall number of active users in a channel.

In addition, from a cost perspective, the level of secondary retention data also determines the difference in secondary retention costs. You can see how big the cost gap is from the figure below (below).

If the second retention cost can be 50%, but you only achieve 10%, the second retention cost difference is 80 yuan per person, which is the wasted cost.

2. Since the access path for new users is relatively simple, retention strategies are easier to implement and more focused.

Most new users use the product by actively opening it, and the usage path is relatively simple. Therefore, you only need to optimize this one path, which will make you more focused and more efficient.

In addition, you can also perform refined operations on users based on channel sources. Some channels can obtain users' demographic attributes (gender, region, interest preferences, etc.), which is the basis for refined decision-making.

3. The main line of new user retention

The main line of user growth is the AARRR model, and the main line of new user retention is the conversion from product launch to key behaviors (as shown below).

1. Product launch

Product launch is the source of users using a product and is the first step. That is, the user downloads and installs it in the application market or Appstore , and then opens the client.

This is the access path for most of them. In addition, there are also those brought by external channels. The external landing page is the details page of the content or product, which can trigger clicks to download, install and activate. After activation, it will still land on the same page, but it is only seen on the terminal.

2. Key Behaviors

For content-based products, the key behavior is content consumption, such as reading; for e-commerce products, the key behavior is conversion to new customers, that is, the completion of the first order.

This key behavior is a bit like an aha moment, such as Facebook's "add 10 friends within 7 days" and Dropbox's "upload the first file". These are all foreign cases. In fact, there are very few people in China who can accurately calculate the aha value because the situation is not so ideal.

But that doesn't mean we can't continue. When it comes to user growth, the most important quality is flexible judgment and quick recommendations, and you cannot rely too much on data . Because there is no timely and complete data for you, and you don’t have time to wait. So as long as you can roughly determine a key behavioral indicator, you can start working on it.

IV. Common Problems in Formulating Indicators

Setting indicators is a very important task in user growth, because you need to use indicators to bring together members of different teams, and it would be best if you can cover everyone's original indicators to achieve a win-win situation. If there are no common indicators, or if everyone does not agree on them, then this project team is basically dead in name only.

There are two elements to setting reasonable goals (pictured above):

1. Decomposed from the overall product and company goals

For example, as mentioned above, assuming that the company’s goal is revenue, this is also the goal of most companies. To achieve this goal, we need to have a sufficient base of active users and the ability to monetize, both of which are indispensable.

A sufficient base of active users is assumed to be DAU. This is the breakdown from overall revenue target to DAU and commercial monetization.

Next, DAU can be broken down into new active users + retention of old users. Therefore, adding new active users is what this article is about.

Let's summarize it:

Take "company revenue" as the North Star indicator.

First level analysis: Company revenue = number of users (DAU) + monetization capability

Second level analysis: Number of users (DAU) = new active users + old user retention

The third level of analysis: New active users = New users * New user retention rate

This article talks about how to improve the " retention rate of new users".

What is unacceptable is that there are no common core indicators and there is no logical connection between the indicators of various roles. If the goal is not broken down clearly, deviations will occur during the specific implementation.

2. Mutual constraints and checks on multiple indicators

The indicators set cannot be single, and multiple indicators are needed to form constraints and checks to determine whether this measure actually brings about growth.

For example, you are responsible for the push module and a single indicator is set for you: "the number of users who open the push function." That's very simple. If you send more push notifications, or more clickbait and vulgar content, this indicator will increase, but user satisfaction and subsequent usage time and retention will decrease. Therefore, there should be multiple push indicators, such as the number of users who open the app, the opening rate, the contribution to DAU, etc.

For another example, you are responsible for a product module and the number of users visiting the product module is set as an indicator for you. Then your focus will become increasing the exposure of this module and applying for more resources, rather than improving the product experience.

Therefore, three indicators for new user retention can be set, namely the number of new users, next-day retention, and customer acquisition cost. Because the number of new users and secondary retention are inversely proportional, if you want to achieve an increase in the latter, you can just reduce the former.

But if the goal is to increase the number of new active users, as the formula above says: New active users = New users * New user retention rate, so both the number of new users and the retention rate need to be increased.

5. How to retain new users

Generally, there are four methods: refinement of user operations, product process optimization, activity incentives, and provision of high-quality services. (Picture below)

According to the main line from initiation to key behavior conversion, the above four methods are actually serial processes.

Step 1: Refine the customer acquisition channels and carry out targeted user operations for different channels. You can also screen channels to obtain high-quality traffic .

Step 2: Convert the acquired traffic into key behaviors more efficiently. It is necessary to optimize the product process and improve the conversion rate at each step.

Step 3: In order to improve the conversion rate at each step, you can also guide users to convert through activity incentives.

Step 4: When users complete key behaviors, the product should provide the best service. Only when users are satisfied will there be subsequent retention.

Here are the four steps in detail:

1. Refined operations

First of all, refined operations are to find and analyze the sources of users, and put the decision-making stage in front, so that we can obtain different characteristics of users from different sources.

For example, information flow channels can be used to carry out targeted delivery according to fixed groups of people. While defining the users, it can also obtain the attributes of the group of people. In this way, we can develop corresponding retention strategies for this group of people. For females + users in first- and second-tier cities + interest points, recommending content or products with good conversion rates can all be a purposeful attempt.

Secondly, you can also carry out refined user operations through activities. Users who are motivated by a certain type of activity can be considered to meet the needs of this type of users.

For example, in an activity of books + movies + music, some users only participate in the reading part, so we can assume that these users are interested in books. After returning to the client, you can recommend content or products related to the book.

For example, a red envelope activity attracts a group of users, and it can be assumed that these users are all profit-seeking. After returning to the client, you can push profit-seeking activities to guide users to continue to stay, as shown in the figure above.

2. Product process optimization

To increase the conversion rate from product launch to key behavior, you need to list each step and see where there is room for improvement in the conversion rate.

For example, why don’t high-traffic products like Toutiao and Baidu Mobile collect users’ interests? By allowing users to fill in their points of interest, recommendations can be more accurate.

As shown in the screenshots of the two pages on the left in the figure below, they are collecting points of interest; while UC and Himalaya on the right have reduced the entrance to collect points of interest. You can set it by clicking it, or you can ignore it so that it does not take up any space on the page.

Logically speaking, every additional step a product takes will result in additional loss. It's just a matter of weighing the benefits of this extra step versus the loss, which is more worthwhile. Judging from the situation of Toutiao today, collecting points of interest is definitely a negative thing.

This is just one case, a step for new users from initiation to key behavior conversion. Only by breaking it down can we know whether there is a problem.

3. Activity incentives

Through activity incentives, new users are guided to perform key behaviors and thus retain them. Take Meituan app as a case analysis, see the figure below.

The operation process for new users is shown in the picture above. The first thing they get is a red envelope for takeout. After receiving it, they return to the Meituan app and see a pop-up window for red envelopes for new customers. After closing the pop-up window, they see a gift package exclusively for new customers. There are activity guides at every step.

In addition, users' original understanding of Meituan is "discounts", so such activities are in line with user expectations.

4. High-quality service

When new users are using your product, you should provide the best service or products and seize this most important opportunity to retain these users. Just like the example of opening a restaurant above, when customers come to dine for the first time, you have to serve the best food to satisfy them so that they will come back for a second visit.

I think this is one of the reasons why restaurant specials are placed first on the menu. Old customers know what to eat without looking at the menu, but new customers don’t know how to order, so the best food is placed at the front of the menu. The same is true for making Internet products.

The three pages in the picture above are Weibo, Zhihu and Tieba in the non-logged-in state. Because the user is not logged in, the product does not know who the user is or what his or her characteristics are, so the most popular and high-quality content in the product is pushed out.

6. Common Problems When Doing User Growth

1. Poor data foundation

A friend said that there is simply not enough data to refer to in the company, so it is impossible to achieve growth. I don’t think so, because the data foundation of most companies is very poor. This is the current situation and we must first accept it.

Secondly, data is only a reference and a tool to measure effectiveness, but it is not a factor that determines the life or death of a product. I often say, if data can give you everything, why do you need us?

If the data foundation is poor, build it up bit by bit. Build a data platform, sort out the caliber, formulate processes, etc. As long as you persist in doing it, there will be obvious improvements in half a year or a year.

This is also something that needs to be done to increase user growth. You can’t expect others to cut the vegetables for you and cook them.

2. Lack of growth methods

First, grasp the main thread. For example, the main line of user growth is AARRR, and the main line of new user retention is the conversion from startup to key behavior. As long as you have a main line, you can break it down layer by layer step by step, solve the problem and respond to it accordingly, and you will naturally find a way to grow.

Secondly, communicate more with your peers, learn from other people’s experiences, and apply them when you are confused.

3. No suitable talent

We don’t have any specialists in user growth, most of them changed careers halfway through. We can screen the suitable ones from products and operations. The essential qualities are product thinking, data analysis ability, strong goal orientation, and strong driving force.

4. Inefficient collaboration within the company

It is normal to find it difficult to promote internally in the early stages of user growth. First of all, it takes time for everyone to understand this matter, and everyone thinks it is just a concept. We need to slowly build a culture of growth to enhance understanding.

Second, user growth is often “doing someone else’s work.” Colleagues may think, I’m doing a good job, why should I listen to your instructions?

Therefore, user growth must be promoted from top to bottom, and the boss must approve of this matter. Otherwise, it will be basically difficult to proceed.

Source: Operation Dog Work Diary