APP user operation: How to prevent users from changing their minds?

Preventing user churn and improving retention and conversion rates are top priorities in operations . Nothing can be accepted by everyone, but operators should do their best to cater to the users' appetite and make their products accepted by the majority of users. But how to prevent user churn? Make users not want to leave?

How to define user churn

User churn is actually very easy to understand. It refers to users who no longer use the product within a period of time. But in fact, different products have different dimensions for measuring user churn, and it is difficult to have a specific definition. The three dimensions for defining user churn mentioned in the book "The Light of Operation" are worth referring to.

1. Lost users are determined based on operational goals
2. Lost users are related to product usage frequency
3. Lost users are positively correlated with user groups

In fact, the three dimensions are easy to understand, for example:

• For apps with high usage frequency, such as WeChat , Weibo, Meitu , video apps, taxi apps, music apps, etc.; these types of app products are used very frequently, and increasing DAU is the top priority of operational KPI. Users of this type of product who have not logged in within half a month/a month can basically be considered as lost.
• For e-commerce apps, defining user churn requires segmenting users, where males and females make completely different decisions when shopping.

From a male perspective: Boys will choose things purposefully and efficiently. If they see something suitable, they will place an order immediately. They usually accumulate many things together and buy them at one time. In this case, the dimension that defines male user churn is not half a month or a week, but may even be longer.

From a female perspective: shopping is their nature. When they have free time, they will go shopping and compare prices from different stores. Therefore, for women, the maintenance period that defines loss may be a week or half a month.

User churn is a long-term and continuous behavior that operators need to treat correctly. User churn is not terrible. Once you find signs of user churn, you need to analyze the reasons for user churn through data and then conduct targeted user recall.

Analyze the reasons for user loss through data

When doing operations, you must use data to speak, and not just make assumptions out of thin air. Through daily data statistics and analysis, we can observe the user activity; once user churn begins to occur, I believe we can get feedback from the data. When there are signs of user loss, you should analyze the users and find out the reasons for the loss. You can start from the following aspects:

(1) What similar behaviors did the user have before he or she churned ? By segmenting the behaviors of lost users, including usage frequency and usage behavior (referring to the stage at which users lost), we can roughly determine the reasons for user loss. for example:

1. New users download the APP and try it out, but a high number of users leave after entering the registration page. This may be due to a problem with the design of the registration page: too much information to fill in? Forced registration? The registration entrance is unstable?
2. E-commerce APP users are relatively active, but when it comes to the payment page, the number of orders that are actually paid is relatively small. Therefore, it can be roughly judged that the user churn on the payment page is high. In a way, it can also be said that the user experience is not good when the payment page is designed: for example, the payment process is complicated and troublesome, the server is unstable, etc.

(2) Distribution of channel sources of lost users: This mainly focuses on the loss and conversion of new users. Most APP user groups are relatively broad, so DSP delivery becomes a better way to attract new users to APP. Of course, different channels have different effects and qualities in attracting traffic . By analyzing the data of lost users, we can make a rough judgment on the quality of the channel.

(3) What are the types of lost users? For example, gender, age group, hobbies, occupation, etc.; in the final analysis, has refined operations been achieved?

Suppose a scenario where an e-commerce app launched a product promotion campaign targeting women, but in the data analysis after the event, it was found that the churn rate was even higher.

Then we can use the data to see what the types of these lost users are. Are they all male or female? If it is a male, then there is no user segmentation and refined operation. Why should activities about women be triggered for men? If it is a female, it may be that the promotional products are not attractive, the promotion is too complicated and difficult for users to operate, or the promotion provides a bad experience.

(4) At the user churn node, what actions did the product take? For example: ① After the product is updated, the user churn rate is very high. Is it because the user experience is poor? Or? ② The core functions of the APP product have changed; ③ The server is unstable, resulting in poor user experience for some users and a drop in impression score.

(5) Other factors:

• APP content updates slowly. For example, if an information APP cannot provide the latest hot content every day, then there is actually no need for this APP to exist and users do not need to use it.
• Poor content quality causes user loss. Regarding the recent events involving big Vs on Wuguoqia and Zhihu, many people have actually reported that the content quality on Zhihu has also declined, so they think it is normal for big Vs to be poached from Zhihu.
• The emergence of competitors: the number of apps in various industries is close to saturation, and the competition is very fierce. Every move of competitors can have an impact on their own apps; I. Wukong Q&A poaches Zhihu big Vs, which may draw away some of the big Vs' fans. Ⅱ For example, in the e-commerce APP war, whichever APP has better product quality, lower prices, and more promotional activities will be more popular.
• The reasons for user churn in different life cycles are actually different, and need to be analyzed and hypothesized based on the data of your own APP.

Image source: Zhihu user Wang Wei’s question: “How to conduct a survey on the causes of user churn?”

How to prevent user churn

Although user churn is a long-term and continuous behavior, the goal of operation is to reduce user churn and improve retention rate. Do a good job in user operations and prevent user loss, instead of waiting until you find that users have lost, and then try to make up for it. To prevent user loss, you should do the following in advance:

Preventing user churn from the product level

(1) When revising a product, you need to communicate with operations

When developing products, many people work in isolation. They always think they understand the users, but often fail when it comes to revise the product! You think your user experience is good, but you find that there is a serious loss of users after the revision, leaving a big pit for operations; therefore, when revising the product, you should communicate with operations to see whether the revised functions and experience are reasonable.

(2) Product revisions should be gradual and cyclical, and major revisions should be made with caution

You must be very cautious when making major product revisions, whether it is an APP or a platform product. Major revisions mean that users have to adapt to your product again, and operations have to cultivate user habits again. The prerequisite for a major revision is to conduct grayscale testing, communicate and discuss with operations, and make good plans and strategies.

The revision should be gradual and cyclical, and the entire product update and revision should be broken down. For example, the revision can be divided into three stages. The first stage should be implemented first, and the user acceptance level should be observed through data. Then a decision on the second stage of revision should be made. Good notifications should be made in the early stages of product updates and iterations, and new functional modules should be released gradually to give users some time.

(3) Pay attention to interface design and APP running speed

The interface design of an APP does not require to be very sophisticated, but it must be clear and adaptable to different screens. A good interface can not only bring visual enjoyment to users, but also help users use your product better. A bad interface will make users think it is a copycat product, and therefore they will not be interested in using it.

Of course, the APP needs to optimize performance, improve running speed, avoid jams, and avoid problems such as APP crashes, memory occupation, slow running, etc. Now there is a serious phenomenon of product homogeneity and the cost of user replacement is relatively low.

Toutiao is a large-scale information APP. It has done very well in recommendation mechanism, page design, column planning, interface adaptation, etc., and the user experience is very good. However, I often hear that our local information APP is heavily advertised, but after downloading it, I found that... (I am speechless, see the picture for yourself)

(Today’s headlines/Red Peach Cake APP)

Preventing user churn from an operational perspective

Why do users churn? In addition to the poor product, users do not feel a sense of superiority when using the APP. To prevent user loss, the most important thing is to improve user retention rate and cultivate user habits. I have written "APP Operation: How to Improve User Retention Rate?", which lists:

①Activity operation to stimulate user desire

Event operation is certainly important, and each event can drive data improvement to a greater or lesser extent. But not all types of activities can improve user stickiness. For example, when an e-commerce app is doing a promotion, many users just complete an order for the sake of the promotion, but once the promotion is over, the users do not take any action. Such users can be called one-time users and may be lost at any time.

② Do push properly

On the surface, Push notifications appear to be unilateral communication between operators and users. However, by segmenting users into smaller groups and making precise push notifications, it appears to be two-way communication between operators and users.

③ Check-in function to improve user stickiness

The sign-in function is very helpful in improving user stickiness. By signing in to obtain points, users can continuously spend money, gradually improve user stickiness, and prevent user churn. Users would feel uncomfortable if they didn’t come every day to collect some points and see if there were any changes to the products.

Event operations, Push notifications, clock-in sign-ins, etc. are helpful in preventing user churn, but the premise is that operations must understand users. User operations must do a good job of user portraits , understand users, grasp user interests, and then group/cluster users for operations, and operate in a targeted manner to make users feel that you understand them.

In addition to these three points, there are actually many other ways to reduce user churn and increase the conversion rate of the user behavior funnel through various operational means. Ultimately, it is about improving the user growth system and incentive mechanism to help users continue to grow.

In conclusion:

User churn is normal, and there are many reasons for churn. No product can be liked by everyone, but operators can intervene in advance through some means to prevent user churn to the greatest extent. Preventing user churn is actually like falling in love, both require understanding each other. Operations understand users and help them use products better!