Shulou(Shulou.com)06/01 Report--

This article mainly shows you "how to use data to improve stock customer stability in the Internet". The content is easy to understand and clear. I hope it can help you solve your doubts. Let me lead you to study and learn this article "how to use data to improve stock customer stability in the Internet".

I. which users do you keep?

The goal of user retention is to improve customer stability and customer income. from this point of view, subdividing the user group can subdivide the customer value and customer stability, establish a customer value-stability model, and find low stable users. High-value users have priority.

The first step is to establish a customer value model.

Using the user data accumulated by operators, using data mining method to establish data model is a common method. The modeling process mainly includes business understanding, data preparation, data exploration, model construction, model evaluation, deployment and application.

Figure 1 Business Modeling process

Starting from business experience, customer contribution and customer loyalty, and the stability of business use are all factors that affect customer value. the business indicators included are: customer historical contribution, current contribution, business volume, customer interaction circle size, proportion of local calls, duration on the network, contract situation, remaining contract time, downtime, arrears and so on.

According to the business index, the Delphi expert scoring method is used to give the index weight, form the customer value index evaluation system, and calculate the score of each customer. Clustering the results of customer scores to form a stratification of high, medium and low value customers. If the customer score shows a normal distribution, it shows that the fitting degree of the model is good.

The second step is to establish a customer stability model

In terms of business, the user stability model mainly includes user basic attributes, business bundling, consumption behavior, business use, consumption trend and so on, including: package type, terminal brand, contract bundling, value-added service use, the number of payments in the last half a year, the downward trend of ARPU and other indicators.

By extracting the data of historical stable users and off-line users as a training set, the important factors affecting customer stability are found, and the rule set of stability prediction is output by decision tree model. users are divided into high-stable, medium-stable and low-stable users.

The stability model is evaluated mainly by coverage rate, hit rate and other indicators. The coverage rate refers to the proportion of the users who actually leave the network in the early warning user list, and the hit rate refers to the proportion of the real offline users in the early warning user list.

The third step is to segment the users and determine the priority of customer retention.

According to the degree of customer value and stability, the user is subdivided to form a subdivided user group, and the customer is subdivided into worrying about customers, caring about customers, rest assured, and attentive customers, as follows:

Figure 2 Segmentation of customer base based on customer value-stability

The priority of customer retention is to worry about customers, to care about customers, to rest assured, and to pay attention to customers. When marketing resources are limited, priority should be given to worried users.

Second, how to improve customer stability, we must first find the reasons for the low customer stability, and then prescribe the right medicine.

There are generally two ways to find the reasons for the low customer stability, one is to collect the reasons for users' leaving the network through interviews, and the other is to find out through the characteristics of the data that although the objective data can not directly reflect the reasons for the loss of users, but the information revealed by the user behavior data is often more real than the subjective data obtained from the survey. There is still a distance from "saying" to "doing".

Due to the high input cost and poor effectiveness of interview research marketing resources, operators are more and more inclined to play the role of data. From the data, it is found that there are obvious clustering characteristics of offline users:

The voice saturation of off-network users is seriously insufficient, and the traffic saturation is seriously insufficient or greatly overflowed, showing polarization in the communication circle of off-network users. The attenuation of off-network users is obvious in 4-9 months, and the highest when they are 2 years old.

Similarly, users with high stability also have obvious clustering characteristics:

The offline rate of users using converged packages and value-added services is significantly lower than that of other users using V-net services and larger communication circles.

Therefore, to improve customer stability, we can start from the following three aspects:

First, use the pain point to optimize the product.

Continuously providing customers with high-quality products and services is the foundation of stock customer operation. The most important thing is to solve the main pain points of users using the product. From the above package saturation, we can see that users are not satisfied with the tariff design, and the serious lack of package saturation can increase operator income in the short term, but do some damage to user loyalty. Improve the saturation of packages through primary and secondary card business and traffic transfer, and recommend package upgrades or traffic packages for oversaturated packages to meet their traffic needs.

Second, find the itch point and improve the user stickiness.

The so-called itch point refers to the products or services that users are very interested in and want to buy after hearing it. For example, when Apple's contract machine was first launched, it promised to enjoy zero yuan for higher monthly consumption, which is an irresistible temptation for fruit fans, such as contract machine, contract package, integration business, points and other sticky business.

It is very effective to recommend contract business for users with no contract and low stability. At the same time, it can also superimpose network age growth, points plan and so on to enhance user stickiness.

Third, provide refreshing points for users to stay.

The so-called refreshing point refers to the products that make users feel aha moment when using the products, which exceed users' expectations. The campus virtual network of operators has this power in the first place. By paying a monthly fee of a few yuan, users who connect to the campus virtual network can enjoy unlimited free calls between users within the network, which is still very lethal in the year when there is no Wechat.

Although the charm of the current kinship account and virtual network is greatly reduced, it is still very effective for users with small communication circle and frequent calls. For low-stable users with this feature, users can recommend relevant preferential services that can expand their communication circle.

The above is all the contents of the article "how to use data to improve the stability of stock customers in the Internet". Thank you for reading! I believe we all have a certain understanding, hope to share the content to help you, if you want to learn more knowledge, welcome to follow the industry information channel!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.