How to use data analysis to acquire customers at low cost?

This article is organized as follows:

• How to extend user life cycle ?
• How to save operating costs?
• How to acquire customers at low cost?

Our daily behavior on the Internet generates a large amount of data. With the development of big data and artificial intelligence , the value of this data is becoming increasingly important.

Different development, operation and marketing strategies are needed to match different product development stages. The key is how to control the overall rhythm and strategy of product line development.

Everyone is talking about big data, and here we have distilled several characteristics.

1. The volume of data is extremely huge.

The threshold for data accumulation is getting lower and lower, and people are paying more and more attention to the digitization process from the real world to the data world, and a lot of infrastructure for data collection and data aggregation has been built.

In this process, the volume of our data is getting bigger and bigger.

2. Diverse data types.

Everyone is doing data collection and data analysis , and the types are very diverse.

These data are often not connected to each other. At this time, how to connect the data and how to make the data into a closed loop in one's own field can form a virtuous iteration.

Another thing is that the processing speed must be very fast. If the processed data is not generated instantly, it will lose its commercial value.

How to extract the value from data and release it instantly is also a very important point in user life cycle operations.

3. Low value density.

You will find that in the application of big data in various industries, many times you can only get useful data or the value of the data can only be realized after accumulating a certain amount, and the density is very sparse.

In addition to data, we also need to compete on model algorithms and computing power. If anyone can use better model algorithms and suitable model computing power to connect to data and business value as quickly as possible, this is the key step to success.

I don’t know if you have read Tencent’s financial report. From the user growth data reflected in the report, we can clearly see that the current traffic dividend is indeed gone.

Therefore, with the disappearance of traffic dividends, our user growth has shown a weak state. Why is it weak?

An important factor is that more and more competitors are entering the market, and competition in the entire market is constantly intensifying.

The extensive operation mode is no longer suitable for this market environment.

Current operations require us to be user-centric and deeply explore some of the user's characteristics, including using various data to help us improve the efficiency of user operations .

So we came up with the concept of "user life cycle management".

In fact, when a user enters your APP, whether it is from downloading and installing, to using, or paying, he will disappear one day, and the whole process will go through different stages.

Different stages have different characteristics, including the customer acquisition stage, growth stage, maturity stage, decline stage, and loss stage.

As an operator, how to use data to better serve these users at every stage and improve their experience is actually a question that every user operator will think about.

Next, I will tell you how to use data to do such work in each cycle.

1. Customer acquisition period

In the customer acquisition stage, the means of acquiring customers are similar, which are nothing more than three levels: channels , social media , and advertising.

① Channel management

Channels may be the main means of acquiring customers. Today's APP operators will look for different channels to promote their APPs . In this process, they will encounter some problems, and the cost of acquiring customers is getting higher and higher.

In fact, it is also because of the development of the Internet to this point that many resources have been monopolized, and advertisers’ bargaining power in this regard is very low, so they can only work hard on customer acquisition costs.

During this process, some problems will also be encountered. For example, when advertisers use these channels to carry out promotions , they often encounter the problem of balancing quality and cost.

Each channel will use various means to attract a large number of new customers.

But advertisers often think, when I want quality, the cost may not be guaranteed; when I want quantity, the quality may not be guaranteed.

Therefore, in this process, we need to find a balance point in which data can be used to speak for itself, so that everyone can have corresponding data support in their respective fields.

Advertisers often have some unrealistic requests, such as bringing me 50,000 new customers a day. Everyone knows that the possibility is very small. What will they do as a channel?

He may give you some fake users, users with very low activity levels. At this time, we have to use data to measure them.

What are the characteristics of the users brought to you by each channel, and what are their retention rates and activity levels?

Using these data indicators to allocate channel budgets is also a way to find the balance point using data.

② Social dividend

Everyone knows that social networks are the most advantageous communication tool on the mobile Internet , and everyone will use this communication tool.

Because the conversion rate of old users bringing in new users is actually very high, and social attributes can eliminate most users' concerns.

If you add big data tags, it will enable operators or developers to carry out targeted communications more effectively, which can help you increase user growth.

③Advertising

Advertising is the most common method. People will use various types of advertisements to place advertisements, but in this process, they will encounter a problem?

It is a game between channels and advertisers.

Channels can attract new users through various methods, but sometimes the methods they use are different from what advertisers want.

For example, if I target advertisements at women, most of my new customers should be women. How do I verify this?

When launching an investment, big data is also needed for verification. How to verify it?

At this time, we need to use a third-party data platform. We will use the "Geteng" product under Getui. Through Getui's big data analysis, we can tell you what characteristics this user has and what interests and hobbies he has.

In this process, we will also find a balance point, so during the advertising process, we can use user portraits to improve the efficiency of your operations.

In the customer acquisition process, or in the initial stage, we must use data management platforms for data verification. These data platforms also have different functions in different stages.

2. Growth stage

①Cold start

During the growth stage, everyone will encounter a problem: when a new user comes in, you don’t know what characteristics this user has.

At this time, we can use some third-party data platforms to better understand this user.

It can help you make a cold-start content recommendation, and it will have a very good effect when applied to information flow and e-commerce .

② Content recommendation

Toutiao has always wanted to serve different people in different ways, but this is actually not an easy task.

In my opinion, this is a very slow process, or it is a step-by-step process. Why do I say that?

For example, I make a simple grouping of my users and use different operation strategies for people in different groups, but everyone will feel that this will not have the same effect.

In fact, if I go one level lower, I will do vector-level classification and cluster analysis, put users with common characteristics in one category, do corresponding analysis, and develop corresponding operational strategies for people with these common characteristics.

3. Mature stage

As the business enters the mature stage, operating costs may become higher and higher, but the boss needs to save on operating costs. What should he do?

① Cost savings

I just talked about advertising. It is very likely that the traffic brought to me is fake. How can I tell whether the traffic is real?

We can use a third-party big data platform to help you identify whether these users have installed this APP in recent times.

From which media do these users come, what is their activity level, and whether they are real traffic.

② Cultivate users’ payment habits

For Chinese consumers, the willingness to pay is very low. It is very difficult to get users to pay in China. Why do you say that?

Because Chinese consumers are accustomed to enjoying the services provided by APP for free, they have no intention to pay. So how can we cultivate users' payment habits?

You can start with some products that require small payments to cultivate his willingness to pay.

In this process, if he has gained some different experiences after paying, large payments will not be far away.

③ Preferential incentive policies

For example, sending red envelopes and various incentive methods. Although the methods are cliché, the effects are still very good.

4. Decline

Once a user comes in, it is impossible to say that he will always be your user. He will leave one day. What should you do before he leaves, or when you find that he is about to leave?

In fact, at this time, I think the first thing to do is to do a series of data analysis.

You need to find out whether he is using an APP in the same industry or a competitor's APP while using your APP.

Or whether his needs have changed, these data are unclear to APP developers.

For example, if the user is no longer active today, you will need to use the third-party data operation tool "Number" under Getui to analyze it. It may tell you that the user has installed a similar APP to replace you.

Before the actual loss occurs, you still need to do a series of analysis. For example, the user may be silent, but you still need to judge whether he is a high-value user or a low-value user.

We can judge through big data analysis of his APP usage, opening frequency, and usage time.

After analyzing whether the user is high-value or low-value, a series of awakening actions need to be performed to determine whether the user has uninstalled the app or is just silent.

After the analysis, I also need to determine what means I should use to perform a series of wake-up actions for these users.

5. Churn period

We just talked about churn. Let’s look one level lower. What if the user uninstalls the app?

After uninstalling, everyone will definitely want to recall it.

In fact, everyone knows that recalling users is a very difficult task. Why do I say that?

If we compare recalling users to winning back an ex, it will be easier for everyone to understand.

But before doing the recall, we can do a few things. First, I need to know how many users have been lost.

We can use the "number" of Getui's application statistics products to understand the number of APP uninstalls and uninstall trends in various channels. We can also analyze the user's uninstall flow, uninstall user composition, and uninstall recall analysis.

It’s not enough to just know how many users uninstalled the app. I also need to know what kind of people these users are and ask them why they uninstalled the app.

During this process, we can use data to understand the users of the APP, including which users have uninstalled the app, how many people have uninstalled the app, and the reasons for their uninstallation. In this way, we can help you conduct a recall.

At the same time, you must have a certain understanding of the industry and observe whether the entire industry has changed, and whether it is your own fault, an operational problem, or a product or service problem.

Source: Noteman