Fission is complete, how to do refined user retention work

At the end of the previous article "There seem to be many fission routines, but in summary there is only one", it was mentioned that fission is just the beginning of growth. What is more important is to focus on how to retain users after fission? Many companies or operators may focus too much on growth indicators, but in the United States, this is just a vanity metric. Why? Because whether it is the growth of traffic or the growth of users, this number will only increase and not decrease. As long as there is promotion , the number of users will increase.

Pay more attention to retention, and you will get more real growth

This is the data shared by Professor Zhang Ximeng at a conference. The retention rate on the left is 5%, and the retention rate on the right is 10%. The difference of only 5% in retention rate has doubled the revenue of these two companies in one year, and the long-term value of a single user has differed by 6-7 times.

The current situation in most companies is: one colleague is responsible for user operations , mainly doing new customer acquisition work; that colleague is responsible for event operations , only doing various store activities; and another colleague is responsible for content writing, producing various communication content based on the product. Everyone is responsible for their own work on a daily basis, and only occasionally when working on large-scale projects such as Double Eleven or press conferences will a special team be set up to work together.

In operations work, if you can understand the core essence of operations and string together various operations methods, you can make your operations more systematic and refined. This article will discuss how to carry out refined operations and achieve user retention after fission from the perspective of tag construction and application.

1. Understand that the core of business is users, value, and profit

2. Use tags to create user portraits

3. Formulate a label operation strategy

4. What problems will you encounter when using tag operation strategies?

1. The core of business

Every operator should understand that the three core elements of all business models are users, value and profit.

Simply put, it is the value that can be brought to users and the profit that can be brought to oneself. From the user's perspective, why should they spend money and time with you? What value can you bring to me?

First is the user

Who are your users? How old is he? Where do you live? How much money do you have? How is the family? What are your hobbies? What do you like to eat? What do you like to play? What are your usual shopping habits? If you don’t understand your users’ situation, how can your product serve them well?

Then the value

Is this value a necessity of life, a spiritual enjoyment, or an emotional sustenance? Different users have different perceptions, so your operating model will also be different. The same toothbrush helps clean teeth and mouth, but some people choose manual one, which is economical and affordable; some people choose electric toothbrush, which is expensive but more comfortable and enjoyable. This is how different products bring different user experiences and create different values!

2. Use tags to create user portraits

The best way to understand your users is to create user portraits, describe the characteristics of your user group, and extract the commonalities of this group. To put it simply, it is to label the user group.

Generally, user tags can be divided into two categories: explicit tags and invisible tags.

Explicit tags are mainly used to describe the visual features of user groups. Such as the target users ’ age, gender, occupation, region, interests and hobbies, etc.

for example:

Basic attributes: gender, age, permanent residence, native place, mobile phone number, email address, registration channel , etc.;

Social attributes: family attributes, marital status, education level, assets, income, occupation, etc.;

Interests and preferences: photography, sports, food, beauty, clothing, travel, education, etc.

Implicit tags are internal and deep feature descriptions. It includes the user's purpose of product use, user preferences, user needs, product usage scenarios, product usage frequency, etc.

for example:

User behavior : number of daily logins, login duration, views, comments, likes, interest preferences, etc. within 3/7/15/30 days.

If it is an e-commerce industry, there is also user consumption, the consumption amount, number of consumptions, consumption frequency, first (last) consumption time, etc. in a fixed time period.

Label division can also be divided from three dimensions: people, goods, and places. For example, the figure below is the user label system of an e-commerce company. The specific division should be based on your business needs.

As an operator, you still need to know how to obtain user information and how technical guys label users. If you are not clear, you can simply understand:

Step 1: Operations proposes the required label requirements

Step 2: The technical brother embeds SDK data in the corresponding location on the website or APP to record the user's behavior path, etc.

Step 3: Categorize and manage these data tags and create user profiles.

There are many tracking tools on the market. Traditional tracking tools require tracking one page at a time, which is very troublesome. Currently, there are many tools that can realize non-point-embedding user behavior analysis , that is, no point-embedding is required, and full user behavior analysis is performed in real time. Like Teacher Zhang Ximeng's growing io; Teacher Shi Wenlu's Aladdin can realize the behavior analysis of mini programs without embedding points; if it is WeChat , as long as you access the developed API interface, you can obtain the user's openID and other basic information, such as WeChat name, phone number, gender, region, etc.

3. Formulate a label operation strategy

With labels, you can deliver messages in a targeted manner. Each user will receive different information from you based on the tags you have noted, thus achieving precise operations.

• Accurate content delivery

For example, I am operating a WeChat public account about food. When fans follow our public account , the public account immediately obtains the user's geographic location and then tags the fan with his area. In this way, when pushing articles, different content can be pushed to people in different regions, achieving accurate push to users.

Similarly, when pushing apps, you can also push them accurately based on user attributes.

After Didi obtained the destination information of my frequent taxi rides, it recommended a nearby shared car station to me, which is only 100 meters away from my home.

• A/B Testing

Friends who have worked in new media must have experienced this: they have thought of several titles for an article and were very confused about which one to use.

At this time, you can select two tags with fewer users, and then push them to calculate which article title has a higher opening rate in the same period of time. Then other fans’ push notifications can use this title to increase the open rate.

Similarly, when displaying an APP or other website, in order to increase the click-through rate of the page, you can use tags to perform A/B testing and select the page with the highest click-through rate for full network display.

Before carrying out the fission activity, two fission posters were designed. After qualitative small-scale release, the fission index of the two posters was tested, and the poster that was easier to fission was selected for dissemination.

• Super User Management

In his 2018 New Year's Eve speech, Luo Pang mentioned the concept of "super users". As traffic becomes more and more expensive today, we must shift from "traffic thinking" to " super user thinking ". In fact, there has been a "1-9-90" model in the marketing industry for a long time. Find 1% of super users, 9% of which are important, and use these 10% of core users to influence the remaining 90% of ordinary users.

Using tags is a great way to find and target these super users. First, clearly define your super users, key users, ordinary users, novice users, etc. For example, super users are paying users ; key users are continuously active within 1 month, ordinary users are active 3 times or more within 2 weeks; novice users are logging in for the first time. Then ask your programmer brother to help you label these users accordingly. Users under these labels are managed and operated in a hierarchical manner, encouraging each layer of users to continuously move up to become super users.

4. Overall Strategy

When formulating a tag operation strategy, remember to continuously test and optimize the entire tag operation strategy. At the beginning, you should first propose label requirements based on your understanding of the business, and then initially formulate a label operation strategy. During the operation process, we continuously optimize strategies until the entire business process is fully optimized. The process can be summarized as follows: setting standards, screening users, formulating strategies, boldly trial and error, revising strategies, continuing trial and error...in a continuous cycle.

For example, when I was working on the official account of a snack brand , I needed to test which QR code format and copy would make fans more willing to scan the code and follow, so I made three versions of the cards for package delivery.

I asked the technicians to help me set up three QR codes with parameters. Fans only need to scan the QR code of a certain card, and the program will automatically label the fan with this card. After 2 weeks of use, you will know clearly which of the three versions of cards is the most effective. Later, we will push refined qualitative content based on labels such as user origin, snack preferences, and region (some regions prefer sweet food, some regions prefer spicy food).

Some new tags

• Difficulties in user operations

After reading this article, you may understand some of the core elements and operational methods of user operations, but if you are asked to do it right now, you may not know where to start. There are all kinds of problems that you simply cannot solve on your own.

1. To achieve refined user operations, technical help is needed to obtain user data, embed points, label, and other behaviors, and background support is required.

2. This system label gameplay does not have a complete operating model. It needs to be implemented based on the actual business and company conditions. It requires colleagues who are very clear about the business, understand the company's current situation, and are familiar with various operating models to implement it.

3. As much user data as possible is needed to make this mechanism work easily. Otherwise, the amount will be too small and the accuracy will be very low.

4. When there are more and more user tags and operation strategies, you will find that some users will be too disturbed or the tags will overlap multiple times. Then we need to formulate more precise operation strategies.

Of course, companies don’t need to be so pessimistic if they don’t have the technology. There are already many such tools on the market that can assist in operations. For example, I believe everyone is very familiar with Youzan, which can realize various tag functions. Some fission tools have also developed such functions, such as Jipai mentioned in the previous article, which can make qualitative push to each user who comes in through the fission activity after conducting the fission activity.

• Final summary

Being able to truly realize data-driven and refined operations cannot be achieved by just one person or one department. However, as a company operator, you need this perspective and thinking to step by step build an operating system based on user process conversion as the core. Whether you are doing community or other user operations work, you should not only consider the user growth part. The key point is to think more about how to retain users and convert them into your super users after the user growth.

Source: Swimming