How to carry out refined user operations from 0 to 1?

Today, let’s talk about thoughts on user operations .

Generally speaking, user operation aims to maximize user value.

What user operations need to do is actually to continue to pay attention to users

Contact - cognition - experience - use - habit - sharing - addiction

This series of processes will improve new additions, retention , activity, payment, and sharing through various operational means.

Among these, there are two work modules that I believe constitute the core of user operations. One is the user stratification reach system , and the other is the user growth incentive system .

You may have heard a lot about the user growth incentive system. We will discuss this in detail later. Today we will mainly talk about the former.

A typical application scenario is that you are dealing with a product with over one million users, the user growth is slowing down, and the budget and manpower are limited. How can you implement refined management to allocate operational resources efficiently?

For example, in the e-commerce marketing system, there are many kinds of coupons: red envelopes, full-discount coupons, limited-time coupons, points coupons, etc. Who should different promotional coupons be sent to, and how should they be sent to maximize profits?

Throughout the entire process, the role of user segmentation is obvious. We can purposefully formulate more targeted operation strategies - refined operations - based on the differences in user groups at each level.

Its greatest value to operators is to productize operations through refined operations, forming standardized population-strategy-reach-feedback-optimization; making operational resources efficient and controlling each input and output within the most reasonable effective range.

In order to do a good job in this part of user operations, you need to be clear that the process from user activity to conversion is by no means two simple stages. During this period, users must evolve dynamically.

That is to say, we cannot just focus on the user’s current state attributes, but we also need to pay attention to why the user upgrades from state A to state C, and what “steps” need to be given to the user. And what kind of changing trends will occur in users' behavior on the platform in the future.

And as the user's status upgrades/regresses, there must be clues ( user behavior trajectories) that can be traced.

In the specific implementation process, we usually define the status of a certain user group, look for users in this status, and then operate in a targeted manner. That is, the hierarchical model is fixed and the users are fluid.

The user stratification reach system is roughly divided into three aspects. The first is user segmentation, the second is user strategy formulation, and the third is user strategy reach.

There are three steps to user stratification:

Define core business indicators - User stratification dimensions - Establish user stratification standards

1) Define core business indicators.

First of all, defining core business indicators can help us find the key links and factors that affect core goals from the complex data indicators, and guide the subsequent user stratification dimension division.

From a business perspective, current mainstream products can be divided into traffic- based businesses and revenue-based businesses. Traffic-based businesses focus more on user activity (number of active users, active days, and active duration), while revenue-based businesses focus more on paid revenue (conversion funnel, repurchase rate, Arpu value). Of course, the premise of both is user scale.

The AR PU value is more based on factors such as cost and premium, and the operating room for operators is not large. The paid conversion rate requires us to use the conversion funnel. The user scale involves two aspects: increasing revenue and reducing expenditure. The above two data results require step-by-step segmentation of users and behaviors.

2) User stratification (group) dimension division

In simple terms, the dimensional division of user groups is to establish group labels. So how can we label things in a way that makes the division appropriate? Mr. StarYan (ID: MRstaryan) considers users from four aspects : user attributes, user behavior, user status, and user preferences .

User attributes: refers to the basic objective attributes of the user. For example, gender, age, region, occupation, etc. This dimension tells us who he is.

User behavior: refers to the user's usage behavior process, such as downloading, registering and logging in, clicking and browsing, purchasing and placing orders, evaluating and sharing, etc. mentioned above. This dimension tells us what he is doing.

User status: refers to the user's current status on the product. For example, free users, paying users , active users, etc. can be further divided into active users who are active once every 3 days/7 days/15 days according to their activity level, and can be further divided into high-paying/medium-paying/low-paying user groups or annual card/quarterly card/monthly card users according to the payment amount. This dimension tells us his status on the platform.

User preference: refers to the key driving factors of user behavior. Taking online video as an example, users purchase VIP membership for movies and TV shows. Some users will subscribe to the membership when they see content they like. This is content-driven, while some are more sensitive to price and will only purchase when there is a discount or promotion. This is price-driven. The content-driven level can be further divided into action/romance/science fiction, etc. This dimension tells us his preferences.

With the four-quadrant model of user stratification (groups), it can be said that it basically covers the user characteristics we want to define. Next we can extract it as needed.

Here is another RFM model that is often discussed. RFM stands for Rencency (most recent consumption), Frequency (consumption frequency), and Monetary (consumption amount). The RFM model selects the three factors that are most closely related to paid services to determine the value of users and the corresponding strategies to be adopted.

Its typical applicable objects include e-commerce, and specific strategy examples will be discussed below.

The third aspect: Establishing user stratification (group) standards

We have listed the user dimensions above, and then we use these dimensions to establish user stratification (grouping) standards.

Taking the online video user stratification standard as an example, we extract the users we want based on the three indicators of "user status + user behavior + user preference".

Non-member users: users who have not yet recharged their membership on the platform and users who have recharged their membership but are currently non-members.

Active users: users who have launched the app and played videos for more than 30 seconds in the past 7 days/15 days.

User behavior: Played the TV series "xxx" with the media resource label of ancient costume, and watched it to episode 11 (members who did not catch it can watch episode 12).

The user objects defined by the above indicators are non-member users who are active on the platform (played within 7 days), watched the costume drama "xxx", caught up with the latest free episodes, and did not watch the member episodes.

In addition, we have also defined other user groups, ABCD, etc., which will not be elaborated here.

It should be noted that the formulation of user stratification standards varies from product to product. For example, we divide users into active and inactive users based on the activity level in the user status column. Some high-frequency businesses are considered active if they are started every 3 days, while some medium- and low-frequency consumer businesses may be considered active after 10 days. At this time, if the activity level is divided into 1-day lengths and user behavior is split, the result will become complicated. Each group has relatively few users, which is not suitable for testing strategies and cannot solve business needs.

In addition, there is a category of products that target two types of users. They are both producers and consumers of the platform. Two different goals will arise here, one centered around production on the supply side and the other around services on the consumer side, thus forming a two-way user stratification . For example, Weibo, Zhihu, Taobao , etc. But the classification dimensions remain the same.

After completing the division of user groups, the next step is strategy formulation. Different measures are taken for different user levels.

New users: Common strategies include new user benefits and new user guidance;

Ordinary users: solidify user usage habits, deepen user product usage frequency, and stimulate activity

Active users: large-scale marketing activities , member rights guidance, long-term membership packages

Interested users: refined marketing activities, continuous point-to-point stimulation, and member identity building

Paying users: Build a user rights system, enjoy the welfare club, maintain the current status and make early warnings for loss

Taking the RFM model as an example, assuming that (1, 0) represent the (high, low) value of the user, then:

Important value customers (111): They have recently made purchases, with high consumption frequency and amount. They are key maintenance targets. Appropriate core user organization building is necessary, including rights distribution, regular rewards and benefits, etc.

Important retained customers (011): The recent consumption time is long, but the consumption frequency and amount are very high, which means that this is a loyal customer who has not come for a while. Regular EDM, PUSH, private message, SMS, actively keep in touch and interact with the user;

Important development customers (101): They have recently made purchases with high amounts but low frequency, and are low loyalty customers. We strive to keep users active on the platform, optimize products and services, and help users consume their rights and interests regularly.

Important customers to be retained (001): They have made recent purchases a long time ago, their consumption frequency is not high, but their consumption amount is high. They are key users to be expected to churn. It is important to prepare for churn warnings and strategic engagement in advance.

Due to resource constraints, when we can only invest limited resources, we tend to lean towards the core group, because only the core group can contribute the greatest value.

In the hierarchical strategy model of the user life cycle , the general priority order is existing users > high potential users > pre-churn users > churn users.

In this way, only by grasping the core can the data indicators be quickly pulled back to the market.

The last step is strategy contact, which can be simply divided into message contact and rights contact.

First, let’s talk about message reach. Message reach can be simply divided into off-site message reach and on-site message reach.

Off-site reach: such as EDM, SMS, PUSH, public accounts , Weibo, etc.

In-site reach: various in-site promotion positions

Large companies generally make message delivery into an automated triggering system platform. Product operators only need to formulate strategies and hand them over to developers to implement the strategy product backend. When users are in a certain behavior state, the message delivery will be automatically triggered. Then assist in some important manual operations.

The other type is rights and interests access, that is, the distribution of rights and interests according to the user growth system.

For example, the membership levels of Weibo, QQ, and video websites, users of different levels can obtain different rights and interests. This is an automated user grading of products guided by operational strategies. Users who reach the corresponding level will receive additional resources.

Taking QQ as an example, users can get suns corresponding to the growth value by staying online for a long time, which can make them feel proud and gain QQ benefits corresponding to the growth value. Another example is Zhihu. Zhihu distinguishes between big Vs and ordinary users, and provides high-quality users with more traffic support and operational support, including more exposure , distribution or material, honor incentives, etc. The core is to retain these high-value users.

Finally, regarding user segmentation, we need to understand the following points:

1) User stratification will change with different stages of product development.

Rome was not built in a day, and neither was Taobao’s “one thousand faces for one person”. For different products, their life cycle basically goes through the seed stage, growth stage, maturity stage and decline stage. The only difference is that different products may have different life cycle lengths, and this difference in length is often reflected in the granularity of operations.

The advice for small products with a small number of users is to start with a rough approach and then move on to fine-tuning, and don't get hung up on details. During the stage when you should be doing full exposure, don't waste energy on " fine-tuning for 1,000 users ." The focus of work for large products with a large number of users becomes paying comprehensive attention to details. The higher the degree of refinement, the more stable the user system structure. At this time, an improvement of a few tenths of a percent is a very big leap.

2) The user pie should be viewed from both a larger and smaller perspective.

When you are facing 100 users, you may know who they are and what their characteristics are, although none of them pay. When you are facing 1,000 users, you may only remember 200 of them, although they have not paid.

When you are facing 10,000 users, you may only remember 200 of them, but 10 of them have paid. You must have a very clear understanding of what kind of people they are, their consumption capacity, their payment motivation, user behavior on the platform, product satisfaction points, activity stimulation points, etc., in order to improve the overall efficiency of operations.

above. Hope you can have fun with your users.