How to build a user points system?

Points is a magical word and the boss is always thinking about it.

When user retention and activity are not very impressive, the boss will always say: "Hey, our user retention and activity are not high, please design a points mall for me."

When you hear this, you may feel overwhelmed. Building a points system has always been a complicated matter. Because it involves issues such as how to obtain points, how to consume them, and how to maintain them.

The acquisition of points should be linked to the behaviors we hope users will frequently perform in the product, and the consumption of points is related to the cost. If the budget and the value of points are not determined in advance, the project could suffer losses at best, or the company could go bankrupt at worst.

Don't disbelieve it, there are really companies that have gone bankrupt because of poor planning and control of points issuance.

Air Canada failed to make a good financial plan, allowing points to accumulate indefinitely, which eventually led to insolvency and bankruptcy. The company eventually spun off its frequent flyer program into AeroPlan to operate independently.

However, if points are used well, it is indeed a good way to increase user stickiness and promote user activity. Today we will talk about how to build a user points system .

Disclaimer: The following content comes from studying and reading articles. The author has not practiced it himself, so let’s learn together.

Do a good job of point cost planning, don't play yourself to death

For paid and non-paid products, such as Taobao and Qutoutiao, the methods for determining the points budget are different.

Generally speaking, for paid products, the method used is: fixed amount ratio method. The fixed amount refers to sales, turnover or net profit.

Generally, for the sake of insurance, a certain percentage of net profit will be used as the cost of point redemption, for example:

An e-commerce platform has sales of 30 million and a net profit of 10 million. The boss waved his hand and allocated 10% as the points budget, so the point redemption value this year is 1 million.

For non-paid products, the usual approach is to subsidize active users with xxx yuan, where the definition of active users depends on the actual situation of the product. Is it necessary to clearly define the exchange rate between platform points and RMB? In other words, it is clear how much is 1 point worth?

In fact, many platforms do not define it this way. The main reason is: we need to leave room for adjustment of the value of points to prevent costs from getting out of control due to users acquiring too many points or redeeming too many points in the future.

Points acquisition, how to define the point value of user behavior

We all know how to build a points system. In essence, we just use points to encourage users to use our products according to our ideas and take actions that we want them to complete.

Therefore, the points system is a very important part of the user incentive system. Together with the membership system, identity and honor system, it constitutes the user incentive system.

The points system is more of a naked stimulation for users from the perspective of material money. It takes advantage of the user's greedy nature and loss aversion.

To obtain the points system, we must first clarify the core and non-core behaviors of the product.

• Core behaviors: behaviors that have a direct impact on the product’s North Star Metrics, such as ordering behavior in e-commerce, posting or reading behavior in community products
• Non-core behaviors: behaviors that have an indirect impact on the product’s North Star metric, such as registration, commenting, sharing, and attracting new users.

Different behaviors will naturally be assigned different points. When assigning points to core and non-core behaviors, focus on:

1. There must be a clear formula for acquiring core behaviors. For example, in an e-commerce app, order amount*1 means buying a product worth 100 yuan, and you can get 100 yuan in points.
2. The points earned for non-core behaviors are designed based on the importance of the behavior and the frequency of earning it. For example, registration behavior is very important and is a one-time behavior, so more points can be awarded. A mild and daily behavior like signing in can be awarded with fewer points.
3. The focus should be on whether the score ratio between core behaviors and non-core behaviors is reasonable. This cannot be viewed in general terms, but must be judged in conjunction with the user's value.

For example, in an e-commerce app, high-value users have a high consumption frequency, so the ratio of their core behavior points to non-core behavior may be 7:3. For low-value users, the ratio of core behavior to non-core behavior may be 3:7.

After assigning points to all behaviors, how do we measure whether they are reasonable? At this time, we need to calculate the maximum amount of points that users at different levels can obtain every day.

Take e-commerce APP as an example: We can confirm the average annual order amount and the average number of new users of low-value users and high-value users by pulling the order data from the background.

For example: low-value users place an average of 10 orders per year, with a value of 2,000, while high-value users place 100 orders per year, with a value of 40,000; low-value users attract an average of 5 new customers per year, while high-value users attract an average of 10 new customers per year.

The points that a single user at different levels can obtain each year are as follows:

Then according to data statistics, we can know that the total number of users at each level, assuming 8,000 low-value users and 2,000 high-value users, then the total number of points obtained by all users each year is 426,650,000.

Our budget is 1 million per year, so the point value conversion rate is: [Point value conversion rate] = 426650000 / 1000000 = 426.65

Rounding off the decimal point, it is 400:1, which means 400 points are equivalent to 1 yuan. If you feel that this ratio is not reasonable, you can adjust the value of some points according to this table.

For example, if the points redemption ratio must be set to 100:1, then the total points cannot exceed 100,000,000. Then the points adjustment ratio is: [Point adjustment ratio] is: 426,650,000 / 100,000,000 = 4.2665

This ratio can be used to adjust the points earned for each behavior.

But actually, there is a problem here, because the total amount of points we calculate each year is calculated based on the maximum amount of points earned, which is actually greatly exaggerated, because for example, not all users will sign in, comment, and post pictures every time.

If conditions permit, the most scientific approach is to use data analysis to confirm the number of check-ins, comments, and photos posted by low-value and high-value users. The [total annual points for all users] calculated in this way is more reasonable.

Points consumption: how to get users to spend their points

Points consumption is the way users perceive the value of points. Generally speaking, it can be divided into two types: passive consumption and active consumption.

Passive consumption means that points need to have a validity period. If the user does not use the points, the points will automatically expire and be cleared after a certain period of time. Passive consumption generally takes two forms:

1. Fixed-time clearing means clearing all earned points at a fixed time, for example, clearing all unused points from the previous year on January 1st of the year.
2. Relative time reset means that the points will be reset to zero at the same time of the second year, starting from the date when the points are obtained. For example, if the points obtained on October 1, 2019 are 100, then on October 2, 2020, the points will be reset to zero.

Active consumption is generally the points mall, including physical and virtual goods. There will also be some operational activities, such as points draws or other points activities.

So, how do we define the value ratio between points and goods? In other words, how to price the products in the points mall?

Generally speaking, all commodities have a cost. If they are produced by the manufacturer, then there is a production cost. If they are purchased from outside, then there are purchasing and logistics costs.

The goods we usually buy have a certain premium on top of these. The main benefit of setting a premium is to facilitate subsequent discount activities and avoid heavy losses caused by scalpers.

Generally, when goods are put on the shelves of the points mall, there will be a pricing formula: Product price = purchase cost * points to RMB ratio * premium parameter A * premium parameter B * premium parameter C….

The premium parameter A is the logistics cost, and the premium parameter B is the tax rate for financial calculations. If necessary, other premium parameters can be added.

So how do we determine the premium parameters? For example, if you want to put a laptop computer on the shelves, the purchase cost is 3,000 yuan, and the logistics cost is 100 yuan per unit, then the logistics premium parameter is: logistics premium parameter = 100% + 100 / (3000+100) * 100% = 103%.

Assuming the financial tax rate is 5%, the financial tax rate premium parameter is: logistics premium parameter = 100*(1+5%) = 105%

A random premium parameter C = 150% is given to facilitate subsequent discounts. Assuming that the exchange rate between points and RMB is 100:1, then the point pricing of a computer costing 3200 in the mall is: 3000*103%*105%*150% / 0.01= 486,675 points, rounded to an integer of 490,000 points.

With the above premium, when we do promotional activities, we can give a discount of up to 30% on a computer worth 490,000 points, that is, 343,000 points, which can just cover the purchase cost, logistics cost and tax cost.

Points maintenance to keep the points system running normally

After the points system is built and officially launched, this is not the end, but just the beginning.

At the beginning, the points system was only based on our estimates of some data. During the operation process, we need to use some data indicators to ensure the normal operation of the points system. These indicators mainly include:

1. Maximum number of points that can be issued = point cost * point redemption ratio;
2. The actual number of points issued;
3. Points redemption rate;
4. User utilization rate.

The first two indicators can determine how difficult it is for users to obtain points, as well as control the amount of points issued so as not to exceed our budget.

When the maximum number of points that can be issued is less than the actual number of points issued, it means that it is too easy for users to obtain points. Then we can look at the data specifically to see which behavior issues the most points. We can then reduce the points obtained by this behavior accordingly, or increase the difficulty of obtaining points.

When the maximum points that can be issued are much greater than the actual points issued, it means that it is too difficult for users to obtain points. Then, we can use the data to determine which behavior should have its points-obtaining difficulty reduced.

In addition, for each behavior that can earn points, we also need to monitor in real time based on data which behaviors earn the most points and which behaviors earn the least points.

Because our goal is to use points to motivate users to do the actions we want them to complete, when they get fewer points for the actions we most want them to complete, it means that our incentives for users are misaligned and we need to make corresponding adjustments.

The last two indicators are points redemption rate and user usage rate.

The points redemption rate refers to how many of the points issued are used, and the user utilization rate refers to how many users who have earned points use them.

These two indicators reflect users’ awareness of points and the frequency of using points, which can ultimately reflect the effectiveness and operational status of points.

The more points users redeem, the more actions they complete, which means they will promote the product more.

The more users use points, the more users recognize our points mall, which means that the redemption rewards we provide to users are what they need, and the more attractive points and points mall are to users.

To sum up

To sum up briefly, if you want to build a points system, you can go in four steps:

• The first step is to determine the budget for points, which is usually drawn from the company's marketing funds or profits. The points system is the icing on the cake. If the product itself does not reach the PMF state, no matter how good the points system is, it will be useless.
• The second step is the source of points. Sort out the core and non-core behaviors of the product, divide users into categories according to certain rules, and determine the proportion of core and non-core behaviors completed by users at different levels based on historical data, so as to adjust the points assigned to core and non-core behaviors;
• The third step is where the points go. Users must have a place to use the points they earn, otherwise the points will be useless. Just like holding money in your hand but not spending it, it is just a pile of waste paper. The consumption of points is mainly divided into passive and active. The passive feature is that an validity period must be set for the points to prevent them from accumulating indefinitely and causing loss of control. Active consumption means opening a points mall or points activities to provide users with a place to spend;
• Step 4: Points maintenance and evaluation. After the points system is launched, the points data must be checked in real time so that the difficulty of obtaining points and the amount of points for different behaviors can be dynamically adjusted to prevent incentive misalignment and uncontrolled distribution of points. The evaluation of points can be judged based on the points redemption rate and the user's usage rate. These two indicators are the best choice for judging users' recognition of the points system.

The above is the result of my study and summary of articles. I have not had the opportunity to practice it myself, so it is inevitable that there may be some theoretical talk. Welcome to criticize~

Author: Jarvan

Source: Jarvan