How to improve new user retention rate?

I have been busy at work recently, so I reposted some good articles from the accounts of various big guys. If you are interested, you can check out the previous articles. Getting back to the topic, I have been busy with Zengshang’s work recently, and have been trying to find ways to improve the retention of new users . Strike while the iron is hot, today we will talk about some thoughts and issues regarding new users (NU: New User).

Question 1: What is a new user?

When seeing this question, many students may be a little confused. Isn’t the new user just NU? Aren’t users who register on the same day considered new users? !

First of all, there is nothing wrong with this answer. However, adhering to the principle of quantifiability and controllability, let us re-examine and specifically "land" the "new users".

From a process perspective, the “new” state is the journey from downloading the App to registering the App. The users we refer to as users who registered on the same day are users who are at a certain time point in this journey.

Because we still have downloading users, activated users, trial users, and finally registered users (here we only discuss the process of common apps).

In other words, there is nothing wrong with defining new users in a narrow sense as registered users. But if we want to better measure and optimize the entire process, we should classify downloads, activations, trials, and registrations as new users. But this should be called the new user period rather than the new user period.

Correspondingly, what we need to optimize is not just a node, but the process of users transforming from market traffic to product users. What we want to optimize is the entire funnel, discover the nodes with weak conversion from the funnel, and thus improve the user conversion rate of the entire process.

Let's expand it a little bit. Once a user registers and uses the service, does that mean they are no longer a new user? Actually, it’s not.

There is a concept of "user activation" in user growth. In simple terms, it means that when users understand how to play the product for the first time and get the point that the product wants to convey, it is considered activation.

We apply this to the new user period and find that users do not have a good understanding of the product before activation. From a business perspective, users who have not been “activated” should also be counted in the new user period.

To summarize question 1.

New users refer to newly registered users, but new users under this definition are only a process node. We need to pay more attention to the entire process of the new user period rather than a single node. At the same time, our time period also needs to be extended from a certain day to a period of time.

Question 2: How to define the new user period?

As I just said, what we need to consider is improving retention during the new user period. Since it is a "period", it must have a starting node and an ending node. In simple terms, the new user period starts and ends at a certain point in time.

According to the definition of channel, we can judge that from the moment we download the APP from the market, we have entered the new user period. But how do we determine at which point a user ends the new user period and enters the long-term active user period?

We mentioned above, user activation. Only when users get the core functions of the product can they be considered activated. That is, only when the user is familiar with the product can we determine that the new user period is over. In other words, what we need to define more is the time point of user activation (the following will explain in detail how to determine user activation).

When it comes to activation, we have to talk about the optimization perspectives of two products.

1. Products that increase the pass rate and doping

2. Products that enhance their original functions

To give a simple example, for a social product, sending text messages to induce information recall and increasing the payment to purchase the opposite sex's login information belongs to the former, while allowing more suitable opposite sex to be pushed to users is the latter. But what we need to think about and balance here is more. We need to take both long-term satisfaction and short-term recognition into consideration.

Because in this stage of stock growth, we still need to be concerned about the saying "even good wine needs no bush", but in the long run, a product should do its job well.

Duties revolve around a core, just like the branches of a big tree. Demands can diverge, like branches and leaves, and can be numerous, but the trunk of the tree cannot be broken.

Let’s summarize question 2.

The new user period is the period from downloading the APP from the market to user activation. The user does not know enough about the product and has not experienced the core functions of the product, which is a "confused" period. During this stage, we not only need to improve the user conversion rate in the short term, but also need to continuously optimize the core functions of the product so as to continuously improve the user retention rate.

Question 3: How to improve the retention rate of new users?

Before answering this question, let’s think of two scenarios:

Scenario 1: After downloading an APP, everyone should have seen the "one-click login with local phone number" function more or less.

Scenario 2: When using an APP, everyone should have done "tasks" more or less, such as completing tasks in Duoduo Orchard to get water drops

Correspondingly, these two scenarios are our two methods to improve user retention:

1. Try to shorten the conversion process and improve the funnel conversion rate
2. Guide users to use the product and improve their understanding of the product

Regarding the first point, we can sort out the steps in each conversion funnel, find the weak points in the funnel, and optimize them in a more targeted manner. We can add guiding quotes and shorten the conversion cost. This is strongly related to the business, so I won’t go into details here. There is only one suggestion: funnel splitting needs to be implemented on each specific user operation step.

Regarding the second point, we only need to clarify one question: how many times and within how much time does the user need to complete what behavior (user activation node).

Why do we say so? Let's recall Duoduo Orchard. The tasks there clearly tell us how many drops of water we can get if we browse the products for a certain number of seconds that day.

The reason for this is that we have to tell users very clearly what actions they need to complete a certain number of times within a certain period of time to get rewards. The premise is that we define this standard.

As for how to determine how much time, and how many times what behavior, there are three main steps:

1. Through the retention rate of behavior, we can determine which behavior is more important, that is, what behavior
2. Through the user retention rate, we can find the inflection point of the retention rate, that is, how much time (usually how many days)
3. Through the inflection point of the retention rate of the number of behaviors during this period, we can find out how many times

Of course, algorithm experts can also calculate the correlation between behavior and retention rate to directly find the time period, number of times, and behavior. I won’t go into details here.

Let’s summarize question 3.

We can improve the retention rate of new users by increasing the funnel conversion rate and guiding users to complete a number of actions within a fixed time period. But this is only the most basic. More work is needed to calculate link contribution and continuously refine and iterate the optimization. Generally speaking, we can think from three directions: active use, passive consumption, and user asset accumulation (these three points will be discussed in the next issue).

Author: Patrol Cat Says Data

Source: Patrol Cat Data