Advanced Operations: Analysis of the underlying logic of the user incentive system!

Common incentive systems include point systems, growth systems, honor systems, wealth systems, social systems, etc. What are the differences between them in essence?

When making an app, we know that if users interact less with the app, they may leave without experiencing the main value of the app. In this way, users may develop a favorable impression of the product and naturally fail to remain active.

The incentive system helps guide users to complete the actions we want them to complete (including multi-day logins to promote retention), feel the value of the product, and also allows users to create value for the platform and even deepen the emotional connection between users and the platform.

Contents of this article:

1. What are the common incentive systems (brief introduction + pictures)
2. Understand the incentive system from its essence and the underlying logic of the incentive system (logic block diagram + step-by-step explanation)
3. Frequently asked questions
4. Should an incentive system be used?
5. What third-party incentive systems are there on the market?
6. Point system, what kind of valuable goods should be placed, how many points correspond to how much value
7. Finally, the operation after the incentive system is established is also very important

What are the common incentive systems?

Currently the most common ones are: points system, growth system, there will also be honor (medal) system , wealth system, and social system .

There are many names, but these are names imposed on some product functions by the outside world and cannot be fully trusted. It is recommended to understand these concepts based on purpose and usage.

Next I will briefly talk about these names (concepts)

Points system

It is to use points to reward users for completing necessary actions, and ultimately users can redeem points for items of certain value.

When establishing a points mall, please note:

1. Points must have substantial value to users, and the value must be stable and visible;

2. User behaviors of different difficulty levels require different points.

For example, Dianping has the following typical points system.

Growth System

The growth system will record and display the frequency and depth of users’ use of the app. The longer the user stays and the more interactions they have, the higher the user level will be.

The growth system usually relies on a points system or underlying score logic to upgrade step by step. For example, 100 points or less is level one, 101 to 200 points is level two, and different levels correspond to different user privileges.

For example, keep has a typical growth system:

Honor (Medal) System

When users complete certain behaviors or use certain functions a certain frequency, they can obtain corresponding medals, giving them a sense of honor.

The honor system usually appears with the growth system, but it can also appear independently. (I personally think that the honor system is just another form of the growth system at the presentation level, which is more common in UGC, game, and learning apps)

The following are the medal systems of keep and Mafengwo:

Wealth System

A system that uses platform virtual currency to stimulate users to perform specific behaviors. The wealth system can easily be confused with the points system because points are usually also used as virtual currency to redeem items in the points mall.

If we need to strictly distinguish between the two, I believe that a major feature that distinguishes the wealth system from the points system is that the wealth in the wealth system is "rechargeable", "consumable" and even "cashable".

Social system

The social system can be understood as promoting users to return to the app and complete corresponding behaviors through a series of social incentives. Social stimuli can be likes, comments, collections, noun rankings, etc.

Although I do not agree that the social system is an incentive system, it is undeniable that at present, when social apps are very popular and tool apps are facing low retention issues, social networking is indeed a way out that everyone wants to try. For example, Meitu has opened a community section and job-hunting apps have also opened a topic section.

The underlying logic of the incentive system

I have just introduced many incentive systems, but if we break them down one by one, we can find that they are actually based on the same basic logic.

Look at the picture below (the author highlights! The most important part of the whole article is this picture!)

The pyramid in this picture should be read from bottom to top, with the underlying logic at the bottom and the product presentation layer at the top.

Generally speaking, the purpose of an incentive system is to use incentives to stimulate users to complete specific behaviors.

Underlying Logic

When considering the underlying logic, we have the following questions:

1. What are the incentives? What can stimulate users? (What do users need and what can the product provide)
2. What actions do we need users to complete? (What the product needs)

These questions, translated into product operation thinking, are: What value does the product have for users? What value do users place on the product? Based on this, I have roughly sorted out the following two-way values ​​(based on my personal experience, for reference only):

Then, we can refine these two-way values ​​and sort out the corresponding user needs and behaviors. Take English Fluency as an example:

After sorting out user behaviors according to two-way value needs, these behaviors can be graded to facilitate confirmation of the connection logic (second level of the pyramid) and what kind of rewards to take.

The specific form of stratification and grading needs to be confirmed based on the product's own circumstances. For example, for some products, activity and retention are not difficult. As long as users participate in a product process, their continued activity is inevitable: if a user has purchased a course in English Fluency, then it is highly likely that the user will log in to take classes multiple days later.

At this time, logging in every day is not that important, and the level you are at can be lower. For some low-retention apps, daily login becomes more important, and their relative level will be higher.

So how do we rank user behavior? We can list some important factors and rank them, as shown in the following figure:

The following figure roughly divides the user's behavior value into levels according to the flow brought by the user's behavior and the difficulty of the user's behavior (for reference only)

Middle-level logic - connection

After sorting out the behaviors, the next step is the second level of the pyramid, which requires confirming the connection between user behaviors and rewards.

A common way of connection is to use points as the connection . When a user completes a certain behavior, he or she can get points. A certain amount of points can be exchanged for gifts of a certain value, or the points within a certain stage correspond to a growth level, and different levels correspond to different user needs and privileges.

There are also behavior-triggered connections , where you can get a prize by completing a certain behavior. New customers are usually attracted by such behavior rewards.

As for which connection method should be chosen, I personally think there is no definitive answer. If the product does not have R&D support, you can use behavior-based connection to trigger rewards for user behavior. If there is sufficient R&D support, you can create a points system with points as the underlying foundation.

example:

Keep is based on a scoring system. Users can earn points by completing courses or checking in, and then they can move to the next level when they reach a certain score. Different levels have different corresponding prizes.

In addition, it also has behavioral-based stimulation. For example, when a user completes the training for the day, a badge will appear. The user can post the badge to his or her circle of friends, which stimulates the user's desire to show off or demonstrate self-worth.

ps: About rewards

Rewards are usually derived from the added value of the product (i.e. the "platform value to users" part in the above logic diagram), or simple and crude material rewards. Rewards can be stratified by value. The higher the level of the behavior, the more worthy it is to use high-level rewards to stimulate users to complete the corresponding behavior.

Presentation Layer

After completing the selection of connection logic, it is time to move to the presentation layer. What the presentation layer needs to control at the operational level is the display structure and copywriting (name, rules, prize description, etc.). I won’t go into details here. Please see the screenshot below for an example:

After reading this, readers may find that the honor system, growth system, etc. mentioned at the beginning of the article are all presentation layers that use scores as different displays of the underlying logic. The key to choosing which presentation layer is what the user cares about. If the user cares about honor, then give him honor (such as games and learning products); if the user cares about substantive incentives, then provide him with tiered substantive incentives (such as shopping products).

At the same time, we must also see that the construction of the presentation layer is also a difficult task, especially when it comes to the honor system. How to display honor upgrade reminders, honor copywriting and presentation, etc., all directly affect the user's perception of value. (Many apps like to use “You have more than XXX users” to enhance the sense of honor).

Frequently asked questions

Should an incentive system be used?

In a certain operation course, there is a scenario summary that requires an incentive system, which I personally agree with:

1. If the user base is large, such as more than 200,000, it is necessary to use some mechanism to attract users.
2. The business model requires users to stay in the product and perform certain actions, but the users themselves have limited motivation.

So 1. It needs to be large in size, and 2. It requires users to stay continuously and perform certain behaviors.

So, what kind of products can/do not need to choose to have an incentive system?

Under the premise of a certain number of users, the following products can ignore the user incentive system:

Products that users use for a short period of time or very rarely, such as:

• Short-term teaching (such as IELTS Brother. They claim to break up with IELTS and leave after taking the test, and there is no need to come back);
• Recruitment products (such as Lagou.com. It is estimated that users use recruitment products in a concentrated period of time each year);
• Wedding products.

Can incentive systems only be developed independently?

No, there are already some mature third-party systems on the market that can support the incentive system, especially the points system.

For example, Youzan and Duiba; as for whether the platform functions meet the company's needs, the operators need to explore it on their own.

Points system, how many points should correspond to how much money?

The difficulty of this question depends on the complexity of the relationship between platform turnover/revenue and user behavior. I cannot give a definite answer here, but I can provide an idea and hope to trigger a discussion among everyone.

User-paid products:

User login value <= daily average profit/daily average number of logged-in users

The value of a user's behavior <= the average profit after the behavior / the average number of people who behaved in the behavior

In fact, the underlying logic here is that every interaction between users and the app has the possibility of eventually leading to payment, so each of their actions has a certain probability of payment, and the value of their behavior is roughly equal to the overall profit of the behavior x the probability of completing the behavior. In order to control costs, the point reward for this behavior must be less than its behavior value.

Non-user paid products:

This type of product is relatively difficult to produce, excluding those that have not yet found a stable business model and rely solely on traffic and advertising.

For some types of products, the payment end is separated from the user (i.e., the B-end and C-end are distinguished, with the B-end paying), such as recruitment products. Although recruitment products are not suitable for incentive systems, if you really want to do it, you can do it like this:

Find out the price range that the payment end B pays for a certain behavior on the C end, and then the point value of the behavior on the C end must be lower than this range.

Example (recruitment products):

On average, customers spend 100 yuan to hire 5 people, and the average hiring rate is 20%, which means that 25 people sign up. Therefore, the "sign up" behavior on the C-end is worth 4 yuan.

Of course, there may be many limitations in the display, such as the incoherent and non-closed-loop behaviors of the B and C ends, and the huge differences in the conversion of different types of information, etc. Real problems can only be dealt with as they arise. I hope that friends who have read this can discuss with us if they have any ideas.

at last

After establishing the incentive system, like many products, you still need to spend time studying the conversions at each level, guiding and rewarding user behavior based on data. If the incentive system shows good data conversion in the initial stage, you also need to prepare to increase user perception in the app, organize activities, improve entry points, etc.

It is not difficult to build a basic product. After the product is built, operating and adjusting it according to data and actual needs is the most energy-consuming and challenging task for operators. I hope to encourage you all.

Source: Andersen