What does the user operation system of a tens-million-level product look like?

Earlier, the author outlined refined user operations from the perspective of the user life cycle. This is the management of the user life cycle, focusing on indicators such as new additions, churn, activity, and conversions of users in each circle in the life cycle. In fact, it is more inclined towards the scope of user operations .

The previous article "User Stratification Model" "I Say: RFM User Value Model Is No Longer Useful" proposed a novel multi-dimensional user stratification idea. Users can get a tag for every relationship they have with the product. The basis of this tag is the assignment method, that is, giving each valuable action of the user a score. The higher the score, the higher the user value and the greater the contribution to the product.

The reason for proposing such a stratification method is mainly to use the refined stratification of users as the basis, to professionalize, modularize, and even semi-automatically execute the operation methods (especially used in event operations), so as to essentially improve the operational efficiency and ultimately increase the overall revenue of the product.

In general, the above-mentioned parts are the basis of user lean operation and belong to the tactical level. Before we start discussing topics related to "user lean operations", we will focus more on the strategic level and first talk about the "underlying logic of user operations" that we have to talk about.

According to common sense, it is often impossible for a platform to continue to provide subsidies/rewards/resources to users indefinitely. From the perspective of the enterprise, these acts of providing subsidies and rewards to users are merely a form of return to the users.

Knowing this, we can better understand why a series of levels are designed in the product "User Center". In fact, it is to provide different resource subsidies to users with different contributions more accurately, to save waste of resources while giving back, and to provide incentives to users so that their value contribution increases. This is the basis of user lean operations.

What is considered user lean operation? How to understand "lean"?

User operation is different from activity operation. User operation focuses more on user growth, user retention and user activity (DAU/MAU). Event operations focus on monetization conversion and are more inclined towards marketing. They focus on transaction volume and GMV (e-commerce).

Lean - In strategic management, there is a term called "lean management": Lean management originates from lean production, which is a management philosophy derived from the Toyota Production System. (The Toyota Production System is a book published by China Railway Publishing House in 2006. It is the most classic work of Mr. Ohno, former vice president of Toyota Motor Corporation. The book systematically reveals the secret of Toyota's excellence, covering various concepts of production management such as Just In Time (JIT), automation, Kanban method, standard operation, and lean production.)

Lean management has been successfully implemented in production systems and has gradually extended to all management operations of an enterprise. It has also evolved from an initial specific business management method to a strategic management concept. It can maximize the social value of an organization by improving customer satisfaction, reducing costs, improving quality, speeding up processes and improving capital investment.

In the process of user lean operation, we will first stratify or grade the users, and then implement different operation strategies for users at each level. Therefore, the establishment of a user level system is an important manifestation of user lean operation.

Let’s talk about the role of the user center and how to establish it.

1. User Center

The name "User Center" also refers to "Me"/"My"/Personal Center in the APP, which is a concentrated expression of the user's lean operation system.

Generally speaking, the user center of each product covers almost all modules closely related to user operations, such as user privileges, user tasks, user rewards, event benefits, forums and other operation modules with frequent manual intervention, as well as template functions such as system messages, customer service, and settings. The details vary for each product.

The mobile Internet companies are experiencing a period of sluggish traffic growth, which indicates that product users have entered the "stock" era. This does not weaken the importance of attracting new users (this has been elaborated in detail in the previous article "Operation Ecological Closed Loop - Product Life Cycle - Attracting New Users"), but it is necessary to achieve the "cost reduction" in "increasing revenue and reducing expenditure" to avoid abnormal loss of users.

In the upcoming tracks of various industries, an essential part of a product is to establish a user loyalty cultivation system to make users more sticky, spread word-of-mouth more widely, and create more value, forming a virtuous circle. Therefore, user operation will become one of the new competition points, and adopting an operation strategy of "strong operation promotion" will be more helpful in gaining the favor of users.

1) User privileges/user levels

User privileges are generally linked to user levels, which is an intuitive representation of the user’s personal “status”/”rank” in the product, and is also a data-based reflection of the user loyalty system.

When it comes to user levels, let's discuss the following:

In most cases, user levels only increase and never decrease because the platform can provide users with sufficient or even unlimited resource rewards.

However, if we try to extend the growth level of all users on the platform, the vast majority of users should be able to reach the maximum level within a complete life cycle. Thinking from this perspective, some platforms may not be able to provide sufficient resource rewards to all users. In this case, the user level setting can adopt the principle of "survival of the fittest", that is, the user level can be increased or decreased.

For example:

Didi Taxi Membership: There are multiple levels including Gold/Platinum/Diamond/Black Gold. The membership level fluctuates every month depending on the taxi usage. Here we can substitute a scenario, assuming that high-level user A is calling for a taxi service in area A1. It is known that this user enjoys the priority queuing function. When multiple high-level users are calling for a taxi service in area A1 at the same time, it is impossible to meet the priority queuing needs of these high-level users at the same time because there are limited service vehicles nearby.

There are also bank card customer levels divided into VIP bank cards and VIP customer bank cards, and the corresponding customer level will be lowered as the customer's savings and credit card consumption gradually decrease.

The same is true for the growth level of Zhihu's creators. The negative feedback from content quality and content influence will cause the content influence score to decrease.

Zhihu creator growth level

In addition to these user levels where "either you advance or you regress", there are even more examples of "only advancing but not regressing". The level honors in the game platform are unlimited virtual resources, and the e-commerce platform is an unlimited virtual shelf compared to traditional offline stores. Even high-level members will receive monthly subsidies for free postage, and the postage cost is very low, which is basically an unlimited resource.

So what are the differences between these two user level systems in actual operations?

First of all, the purpose of the user level "advance and decline" is to further encourage users to consume, renew and use products more and more frequently through level stimulation. This level design is generally used for products that focus on increasing transactions, conversions, activity, and usage frequency, such as Didi Taxi, QQ membership, bank benefits, etc.

The strategy of “only advancing but not retreating” focuses on retaining users and settling them, which helps to extend the user life cycle to a certain extent. For example, for community products, e-commerce products, and gaming products, the higher the user level, the stronger the sense of honor/scarcity given to the user.

Therefore, when we talk about user lean operations, we are actually talking about the precise allocation/supply between the resources that the platform returns to users and valuable users.

On weekdays, we often see knowledge about user operations such as the significance of user segmentation and the elements of building user levels in some operations courses or online articles, but we rarely see any systematic guidance from the execution level, which is very bad.

Next, let’s talk about how to build a “user lean operation system” starting from a financial product with a million MAU (over 10 million downloads).

The "User Stratification Model" in the previous chapter takes into account three key elements: age group, RF (M) model, and user life cycle. Can the stratification model be used to consider user levels?

Ignore the XXX name for now

The core function of the stratified model is to provide operators with a data-based, customized model for screening users to carry out more refined operations on users and improve operational efficiency. The core function of establishing user levels is to stimulate users' high-frequency consumption/investment or retain users. The functions of the two are quite different.

We can consider a scenario where user X was a loyal fan of a certain product and had been playing for about three years. However, due to some reasons, he has been less active in recent months and is in a weak period. According to common sense, a loyal user who has been playing for so long should already be at a very high level.

Obviously, the above user stratification model stratifies users based on the value of their output. It divides the user's value output into time periods, which may be the user's value output in the last month or the last two months. Therefore, it is not suitable to directly use this hierarchical model as a scale to measure user levels.

Then the question is, how to build user levels? Let's take financial products as an example, which can be divided into the following four points:

A. Analyze user distribution

First of all, we need to roughly divide users into basic levels according to the user distribution ratio of existing platforms. We can also refer to the average level of user distribution ratio in the same industry as a reference for minor adjustments.

Since the core of establishing user levels is to stimulate users to increase their investment frequency, we need to find out the key actions that affect the core factors. For example, we can pour out the data of users with high payment frequency and high number of activities participation to make a user core action distribution map.

User core action distribution map

Analyzing user distribution is a prerequisite for user grading and is intended to provide a guarantee for establishing a more reasonable user stratification system. Through data analysis and comparison, we finally came up with an optimized hierarchical structure pyramid, which is a prototype of user levels.

Hierarchical Pyramid

B. Core Behavior Assignment

In the second step, we use the assignment method to assign values ​​to user key actions. The general principle is to assign different scores to different actions. The classification of user key actions is as follows:

Single assignment action: key actions such as first card binding, activation, and completion of personal data (nickname/avatar/signature) settings.

Investment assignment actions: based on key actions such as first investment, investment amount, frequency, etc.

Regular value-added actions: daily login/check-in, participation in activities, posting, sharing content, attracting new users and other key actions.

C. Classify users

In summary, we mainly carried out a simple division of existing users and determined the key actions for user upgrades. With the first two steps as a foundation, can we now divide users into levels?

Not yet.

We also need to take into account the users' actual performance, such as the time required to upgrade, and use users with high payment frequency and who participate in many activities as a reference to evaluate the time these users use the product.

First, screen those users whose total personal value creation has reached the ideal range set by the product, and then come up with an "ideal time" and a matching "highest level" as a reference standard.

Next, crawl previous operational data (key actions, investment/reinvestment/check-in/posting/new user acquisition, etc.), and list the key action events for users to obtain points in each month/quarter. The purpose is to derive a "highest score" reference value that matches the "highest level".

Then, we can conclude that a primary user will be upgraded to the "highest level" within the "ideal time" when using the product normally.

For example, if a new user actively participates in activities and investments, he can be successfully upgraded to the highest level after one year of use. However, other less diligent users may take even longer.

This may seem a bit complicated, but the purpose of previous data crawling and comparing industry data is to provide some data support for a grading system that is more realistic and more suitable for user use.

In addition, in order to give users a better upgrade experience, we can also further refine the actual user level based on the above. For example, the product I previously operated had 18 levels.

Finally, after sorting out the previous series of data, key assignment actions, time dimensions, etc., you can basically establish a practical hierarchical model based on the "hierarchical architecture pyramid".

D. User Level Rights

After establishing user levels that match the product, we need to make users believe that these levels are valuable and that they can obtain some benefits (whether it is a sense of honor or physical rewards) at their fingertips, so that the meaning of user levels can be realized. (Levels themselves have no value, they only have value when given by our operators/product people.)

In most people's opinion, the value of levels is mostly rewards, but this is actually a cognitive misunderstanding. User level rights include two parts: rights and interests.

We need to effectively release user rights based on product features, such as quick transfers, credit card repayments, free withdrawals, lucky flips, unlimited credit limits, medal crowns, exclusive black cards and other core functions.

In terms of rights and interests, we mainly focus on actual rewards, including virtual or physical objects. Virtual birthday experience money, growth value, exclusive service customer service, etc. Physical objects can be peripheral dolls, brand joint peripherals, growth gift packs, etc.

After observing the user level systems of many products, we found that they can basically provide level benefits based on platform resources. The value of user level only meets physiological needs and safety needs. If it can be improved to meet users' respect needs, social needs, and self-improvement needs, the value of user level will be greater. For example, the credit rating of Alipay and the learning medals/certificates of some knowledge service products are more about satisfying users' social needs and respect needs, and are more conducive to improving user loyalty and promoting product word-of-mouth communication.

E. Q&A

After the above four steps A, B, C, and D, we can basically establish a new user level system. This applies to most products, but some products still have issues with user levels, such as points, level upgrades, and upgrade cycles.

Points and user level are two different systems, which can be simply compared from several aspects:

In terms of rights and interests: levels are directly linked to privileges, and points tend to be short-term benefits.

Operational level: The level is relatively stable, with the main pattern of only increasing and not decreasing (but there are also cases where it can increase or decrease, as mentioned at the beginning); the points are more flexible and can be increased or decreased.

In terms of time: levels are phased benefits, while points focus on stimulating current benefits.

In previous products, there was generally an operating strategy in which the points system and the user level system were run in parallel.

The main manifestations are: clear and key actions are generally used as the basis for level; the remaining actions can be appeased through points. For example, in Taobao's previous rankings, only the purchase amount was used as the basis for the ranking; while evaluation, sharing, etc. used points as a means to stimulate user behavior. (I will not discuss the points system here for now, I will sort it out later.)

2) User tasks

User tasks are key actions for users to gain points, upgrade user levels, and obtain user privileges. That is, the tasks performed by users are directly related to user scores. At this time, we need to complete the data embedding on the path corresponding to the user task, and at the same time verify whether the action can serve as a key action.

General user tasks include single/basic tasks: first-time card binding, activation, completing personal data (nickname/avatar/signature) settings and other key actions.

Core tasks: Based on key actions such as first investment, investment amount, frequency, etc.

Regular tasks: daily login/check-in, participation in activities, posting, sharing content, attracting new users and other key actions.

3) Activity benefits

This section mainly gathers some platform activities that are usually launched online or creative activities at festival nodes, and is usually the section that event operators are responsible for.

The event mainly revolves around core operational indicators such as attracting new customers, promoting activation, converting transactions and brand exposure. We will elaborate on the event operation part in "Part 4: Product Marketing Skills".

4) Rewards Center

The "Reward Center" mainly displays the rewards users receive after completing tasks in real time, lists the tasks completed by users in different categories, and clearly shows the tasks that users have participated in, which can better enhance the user experience.

At the same time, the reward center can also be a hidden function that encourages users to do more tasks and obtain more benefits.

5) User Forum

User forums have actually evolved from traditional forum posting to "community"/"discovery page", so the user forum section will be expanded in detail in the following "community and discovery page" section.

2. Super User

A few years ago, there was a saying in the circle that "user operation must evolve from traffic thinking to user thinking to super user thinking."

What is a superuser?

In fact, there is no strict conclusion in the industry, but super users must have the characteristics of high activity, high value output and high loyalty.

Drucker said: The mission of a business is to create and retain customers.

Paul Graham, the entrepreneurial godfather of Silicon Valley, once said: The value of 100 super users is far greater than that of 10,000 ordinary users.

So how do you create a super user?

In fact, it is just like a boy pursuing a girl. First of all, the first impression is definitely the most important. Only by leaving a good impression on the girl can there be a chance of further development. Every subsequent meeting can not be careless. After all, girls are always evaluating whether you are considerate, humorous, and have personality. Therefore, you must ensure that this process is particularly pleasant and satisfying, so as to win her heart step by step.

Cultivating super users is similar in nature. The core is to continuously gain a deeper understanding of users, continuously optimize product services, provide more refined tiered operations, and improve the overall user experience.

Super users don’t just exist in concept; they can be found in actual operational work. For example, in the above user stratification model, users with scores between 64 and 125 (in the red block) are super users; and the level 17 and level 18 users above are also super users.

Regarding the operational experience of super users, the author once sold 20,000 packs of products in 20 minutes in a WeChat mall with only about 3,000 super users, bringing in a turnover of 1.2 million and an ARUP value of 400 (food supplement products, about 60 yuan per pack). The effect is significantly better than that of colleagues in e-commerce operations who placed advertisements on JD.com’s homepage during the same period.

It should be further explained that the core of this marketing campaign does not lie in the price advantage of the product, but in the dual effects of user recognition of the product and marketing techniques (limited-time and limited-quantity group purchases, bundled sales, random prizes, and other incentive techniques).

Author: Shengjie

Source: Shengjie