5 user growth strategies from attracting new users to awakening them!

I will share my data operation experience in the life cycle from the stages of attracting new customers, novices, growth, silence, and churn .

01 Attracting new customers: How to improve ROI

The three core issues in attracting new customers: cheating, traffic hijacking, and ultimately low ROI.

1) Inflate the volume.

There are two ways to brush the volume: organic brushing and human brushing. The low-level one is to use a simulator, and the advanced one is to find a mobile phone and simulate the normal download of a person through the mobile phone program. Meat brushing is like group control, where one group manages many groups and uses tasks to make many people download the App. How to identify this method?

First, we need to use data monitoring. AppTrack has a built-in anti-cheating model and a blacklist, which can monitor real-time changes in IP location and battery power.

The second is through portrait recognition. Because the machines do not have portraits, when we look at mobile phone models, we find that many companies use low-end models or those that are about to be eliminated. If you find a large number of mobile phones from two or three years ago, you can determine that they are flashed.

Meat brushing is difficult to identify because, after all, it is done by real people. We can look at the behavior. A few days ago, a friend who works in education was doing CPA promotion and found that this group of users had a very short usage time. Analysis found that this group of users was different from the naturally growing users, and the core functional modules they clicked were also different. This group of users can be considered to be fake users.

2) Traffic hijacking.

First, when a user clicks on an information flow ad, a box will pop up saying "This user has not passed the green review." This approach is good from the manufacturer's perspective. For the safety and stability of the phone, when the hijacking volume reaches 30%-50%, two results will occur. First, the channel effect and cost of the information volume cannot be evaluated, and at the same time, the hijacked traffic of the information flow is retracted.

Based on this background, we launched the traffic hijacking solution. When delivering information flow ads, the two products, U-App AI version and Apptrack, can monitor the entire process from user click to activation. In this link, we can know which traffic is hijacked and to which channel it is hijacked. After tracing back and analyzing the amount of hijacked traffic, we need to attribute this part of the hijacked traffic to the information flow channel and accurately evaluate the cost and effect of attracting new users in the information flow.

3) How to improve ROI: appropriate channel × appropriate creativity = best ROI.

In fact, it is all about finding matching channels and matching ideas.

Let me share two cases.

First is a ringtone app. They first used portraits to see which media active users appeared on and found that most of them appeared in short videos, including Kuaishou and Douyin. Then they made a short video - setting an iPhone ringtone in 10 seconds. After uploading to Kuaishou and Douyin, I found that Kuaishou had the best interaction, so I later did matrix marketing.

The data results show that with a promotion cost of only 5,000 yuan, it attracted 500,000 new iOS users, which is extremely cost-effective!

The second is Qingting FM.

They found that a large number of users are women, and most of them pay attention to maternal and child and education channels, that is, the maternal and child group. They adjusted their strategy, increased maternal and infant product promotions, and reduced the cost of acquiring new customers by 13%.

As mentioned earlier, we can optimize channels and improve materials based on portraits . We also provide portraits when we invest in information flows. What’s the difference?

First of all, Umeng+ has been in operation for 9 years this year, covering 1.65 million apps and having full-domain portrait capabilities. You can do a global portrait analysis of the App, but what if the users you analyze do not have very obvious characteristics? We have intelligent new user acquisition capabilities, and based on global portraits, we help Apps gain insights across the entire network and find precise user groups.

The process of intelligently attracting new users: U-App AI version will identify highly sticky active users, find similar groups of people across the entire network, and synchronize them to the core channels of core media, such as Guangdiantong, Toutiao, UC, and Baidu. We monitor exposure and clicks through the App and then feed back to the group, continuously reducing the cost of attracting new users. Some developers may not have used the U-AppAI version yet. We support providing positive and negative samples of the population and outputting population packages that meet the App requirements through algorithms.

Application case: A live broadcast app reduced activation costs by 29% through intelligent new user acquisition.

02 In the novice stage, how can we help users quickly find the highlights of the product?

There will be a large amount of churn in the novice stage. Within three days of downloading the App, 70% of users will churn (data from AdRoll analysis and research).

Why is there loss? Maybe you can’t find what you need, or the core functionality isn’t working well. In addition, the development of the entire Internet has led to the fragmentation of everyone's time. Users' concentration time is only 8 seconds (data from Microsoft Research). For example, if there is a lag when downloading an app, they will give up.

So how do we do it?

1) Quickly highlight the core value of the product. For example, after downloading the Aihuishou App, the current market price of your phone will be immediately displayed. Like the Du App, on the login page, the background will scroll and play the entire product content, making users feel that this is the App they want.

2) Develop conversion analysis based on core business nodes to understand where users are lost. Some apps believe that registration is so important that they can build a registration funnel.

Taking shared bicycles as an example, there are startup, registration, deposit payment, and first trip. The most important link is from registration to deposit payment. It may be found that 60 people registered, 24 people paid the deposit, and 34 of them left. At this time, we need to circle out this group of people and push them in time to guide them to the next step.

At this time, if the user has paid a deposit, he or she will enter the growth stage, which is the most core stage, because the scale, quality and payment ability of the growth stage are related to the profitability of the enterprise.

03 How to keep loyal users during the growth stage?

To improve retention, we need to find the factors that affect retention. When Linkedin was doing retention analysis, it found that the retention rate of users who added 5 contacts in a week was much higher than those who did not add contacts, so they began to remind users to add contacts during the registration stage. Twitter found in its early stages of development that the retention rate of users who followed more than 30 people at the same time was high, so it made product improvements. When a user comes in, the first thing he is reminded of is to follow people, and the overall retention rate increases exponentially.

Look for core features related to retention. The first group is high-retention users, and the second is low-retention users. We analyzed the usage of functional modules by high-retention users and found that the retention rate of Function A was as high as 80%. We speculated that Function A may be one of the core factors affecting retention. Through experiments, we found two small channels and corresponding versions for testing. We found that function A was one of our influencing factors, so we iterated the entire version accordingly.

This mainly involves product improvement. How to improve retention in operations? We researched hundreds of apps and found three aspects: product, value, and data.

1) For example, in terms of products like Ant Forest, people are more concerned about the fact that 100 million trees have been planted. Ant Forest allows users to open the app four times a day.

2) In terms of value, through points and physical incentives, guide users to open (App) and retain. For example, Fan Deng’s reading activity is buy one get one free. Some people think it is a good deal to spend one year’s money to get two years’ use. In fact, from the platform’s perspective, it is to gain two years of user retention.

3) In terms of data, for example, Taobao’s personalization and Toutiao’s intelligent recommendations. Umeng+ will also release its portrait and data capabilities to help developers perform personalized operations.

04 In the silent stage, how to accurately wake up silent users?

There are quite a few apps whose daily active users seem to be very good, but in fact they have a large number of silent users and have not been effectively awakened. There are many factors that influence silent users, and it is difficult to determine why users become silent. What we need to do is to guide silent users towards the growth stage as much as possible and avoid loss.

It is difficult to awaken all silent users, but we can find high-value groups that are easy to awaken through event analysis, funnel analysis, path analysis, etc., and reach them through designed copywriting and incentives, and through push and SMS. For some apps, the value of a single user is relatively high. We can also use Umeng+’s portrait analysis to find out which media silent users appear on, and attract their activity by placing corresponding advertisements.

There are also some tips. Through version updates, you may find that a large number of users follow the version update but do not use it. At this time, when updating the version, you must state the content of the update, try to attract users to open it, and try to make them think that you are always changing. This is also a way to awaken silent users.

Umeng+ user life cycle management has a silence prediction. There are three ways to wake up silent users and prevent them from becoming silent before they become silent. We have a prediction model for "users with high risk of silence". You can operate in advance and effectively avoid silence.

For example, DouMi first found high-quality people, then stratified the people, and then tested different types of copywriting + incentives, constantly experimenting and iterating methods. As a result, their conversion rate increased by 50%.

05 In the churn stage, is recall the key?

When it comes to churn, many developers think of recall. In fact, the recall cost is far greater than the cost of attracting new customers. If not a recall, what else can we do?

We recommend iterating the product through version analysis, behavior analysis, and portrait analysis. The value of lost users is to better serve existing active users. Also throughout the entire life cycle, Umeng+ will provide a "high-risk churn user prediction model" to avoid churn before it occurs, and will also perform corresponding operations.

Some students may ask, how do we know this group of people if they have been lost? Umeng+ provides "uninstall statistics". By cross-analyzing the two products, Umeng+ Push and U-App, we can find out which user groups have uninstalled the App, and then make a more detailed analysis of these users.

Every time I talk to developers, I ask if Umeng+ is also doing Push? For example, we have Push in the novice, growth, and churn stages. This year, we launched the paid version of Push. Compared with the free version, the push speed has been greatly improved, and there is no limit on the number of pushes. We also provide uninstall statistics, which can save people who uninstall as a group, so that we can conduct more systematic data analysis. For custom user tags, when we push tags, the free version is T+1, which means that you can use the tag tomorrow after you set it today, but the paid version is for you to set it today and use it today.

Message receipts are easy to view and analyze, so you can effectively see the results.

We manage apps in real time through API and support more manufacturer channels. This is also misunderstood by many people. We are not just a third-party push service. We also provide push services to Huawei, Xiaomi, Meizu, as well as OPPO and vivo, which will be launched in early July. This way, we can manage push services from multiple mobile phone manufacturers using a Umeng+ push background.

To sum up, what can App developers do throughout the entire user life cycle? Improve ROI when attracting new users, let users quickly experience the product highlights in the novice stage, keep loyal users in the growth stage, accurately awaken silent users in the silent stage, and provide long-term value and rapid version iteration in the churn stage.

Therefore, growth that is data-driven and based on the user life cycle is truly effective growth.

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 make use of private domain traffic?

Author: Umeng + Fu Wenguo

Source: Umeng Global Data