User operation: How to improve user retention/repurchase rate?

1. The necessity of strengthening retention/repurchase

In recent years, the prices of advertising channels have been rising year by year. With the end of the Internet user dividend, traffic costs have become increasingly expensive. In this context, how to retain every user brought in by the marketing department and continuously generate conversions, as well as tap into and enhance the value of users throughout their life cycle, has become an essential skill for the product.

E-commerce conversion formula: sales = traffic × conversion rate × average order value × repurchase rate

The most direct impact of retention is traffic cost, which ultimately affects repurchase rate.

Retention does not directly equal repeat purchase, but retention is the basis of repeat purchase.

If the cost of attracting a user is 150 yuan, after acquiring this user, there will be profit only if the user makes a purchase and the profit after deducting the cost is greater than 100 yuan. During the entire life cycle of users, improving retention rate and repurchase rate means improving the profitability of the product.

In other words, if the users you attract are lost before the cost of attracting them is covered, the lost users will directly amount to a loss.

Philip Kotler, the father of marketing, found in his research that the cost of acquiring new customers is five times that of retaining existing customers. If the customer churn rate is reduced by 5%, profits will increase by more than 25%. The importance of customer retention is recognized and acknowledged by all companies.

Two core indicators: retention and repurchase.

2. How to improve user retention and repurchase

To improve user retention, we must first understand the two indicators of retention and repurchase.

1. What is retention and repurchase?

Retention calculation

Retention can be calculated based on various dimensions. For example, new and old users can be divided into new user retention and old user retention. Users coming from different channels can also affect the final retention effect.

Retention rate = number of logged-in users / number of new users * 100% (the statistical period can be day, week, month, quarter)

• Number of new users: the number of new users who log into the app during a certain period of time (usually the first full day);
• Number of logged-in users: the number of users who have logged in at least once since logging into the application until the current time;
• Number of active users: number of logged-in users - number of new users;
• Next-day retention rate: (number of new users who logged in on the second day) / total number of new users on the first day;
• Day 3 retention rate: (number of new users on the first day who logged in on the third day) / total number of new users on the first day;
• Retention rate on the 7th day: (number of new users on the first day who logged in on the 7th day) / total number of new users on the first day;
• Retention rate on day 30: (number of new users on the first day who logged in on the 30th day) / total number of new users on the first day;
• Channel user retention: Understand the retention rate of users from different channels divided by time.

Calculation of repeat purchases

Repurchase rate = number of repeated purchases / total number of purchases * 100% (the statistical period can be weekly, monthly, or quarterly)

• Repeat purchases: Some people who made a purchase made a repeat purchase again.
• Total purchases: the number of users who generated orders;
• Next month repurchase rate: users who placed orders last month / users who placed orders last month and continue to place orders this month;
• Repurchase rate in the next two months: users who placed orders in the first two months of the month / users who placed orders in the first two months of the month and continued to place orders in the current month;
• Second quarter repurchase rate: users who placed orders in the previous quarter/users who placed orders in the previous quarter and continue to place orders in this quarter;
• X-order user repurchase rate: users who have purchased X orders/users who continue to purchase;

Negative factors affecting retention/repurchase

The following are some common reasons for low user retention and repurchase:

(1) Product quality: For example, problems with clothing quality include incorrect sizes, raw materials, color fading, loose threads, etc. If customer loss is caused by quality issues, it will be difficult to regain the customer's trust later.

(2) [Service Quality] Damaged express packaging, disputes over returns and exchanges, delayed delivery, long logistics delivery time, inadequate delivery, uninformed hidden costs, overselling, etc. can all be attributed to service loopholes.

(3) [Product Experience] Product design, lag, long H5 page loading time, crashes, too few reviews, incomplete display of product parameters, etc.

(4) [Sustainability of discounts] If discounts are not sustainable, some price-sensitive users will be driven away.

(5) Whether the product meets user expectations: Users expect that you will have product xx for sale, but upon opening the product they find that you do not have it.

(6) [Speed ​​of new product launch] Rapid new product launch can increase the frequency of consumers visiting stores. If products are regarded as content, users will only stay and repurchase if the content is rich enough and constantly updated. Otherwise, users will be lost.

But in reality, various reasons will have a cross-impact on retention and repurchase, which requires detailed investigation and optimization.

How to locate the problem

The following are some ways to locate specific problems:

(1) Customer Service Feedback: When there are issues with product quality or service quality, customer service staff will often get first-hand feedback from users. Good information exchange between product managers and customer service staff helps to identify and resolve problems in a timely manner. If the problem cannot be clearly identified based on user feedback, you can also conduct a follow-up visit to the user based on the user feedback and conduct data analysis on the corresponding problem nodes.

(2) [Survey Questionnaire] Use questionnaires to investigate users’ price sensitivity. For sensitive users, continuous promotional activities can be used to cultivate user habits.

(3) [Product Trial] Product experience issues can be cross-checked through questionnaires and trials, especially by the product manager himself. Only after multiple trials can the product manager break away from the limitations of his original design and discover problems.

(4) [Data investigation] Data analysis and investigation. For example, if you find that the package receipt rate is only 80% (common in cross-border e-commerce), you can analyze the composition of this 20% based on the corresponding logistics data. Which logistics channel has the highest proportion? Which country has the highest share? By checking the logistics time, logistics information delivery efficiency, logistics status updates, etc., you will definitely be able to find a way to increase the package receipt rate.

1. Membership system and points system

When it comes to improving retention and repurchase, the membership system and the points system are unavoidable.

Membership System

The membership system refers to distinguishing user types, dividing ordinary users into member users, and then providing different products and services or differences in discounts for ordinary users and member users. By managing users through a membership system, we can more effectively strengthen the sense of belonging of these users and even obtain more user data/information, thereby understanding their interests and consumption habits and exploring their intended needs.

(1) What products are suitable for building a membership system?

Not all products are suitable for building a membership system, or it is not appropriate to build a membership system in the early stages of the product when a foundation of user trust has not been established.

1) Users already have a certain understanding of the product. The membership system is based on a certain level of user retention. If user retention is very poor, it is definitely not because the user lacks a membership. Members can only reduce the chance of user loss through their membership on the basis of a certain retention rate.

2) It has the characteristics of high frequency demand. For example, fast-moving consumer goods such as vegetables, rice, oil and salt are necessities of life, and users will be happy to obtain membership to obtain discounts and accumulate points. But if it is a home furnishing product, users’ willingness to obtain membership will be much weaker. After all, you don’t know when you will come back next time.

3) Able to build membership benefits that are attractive to users.

4) Capable of fine-grained member operations. Member users are high-value users and will have much greater output than ordinary users. On the basis of encouraging more users to become members, we must also carry out refined operations for users who have already become members.

(2) Membership types can be divided into the following categories:

1 ) Undifferentiated membership: Undifferentiated membership can also be called universal membership. Users can obtain membership by providing relevant information, and there is no difference in membership rights and interests.

2) Level membership: Different levels are divided according to user behavior data, and members of different levels enjoy different rights and interests.

3) Paid members: Give paid users special benefits.

In a product, several types of members can appear separately or at the same time, depending on the extent to which users need to be grouped in order to maximize benefits.

(3) Which types of users can become “members”

The criteria for selecting users to become members vary according to different membership types:

1) Undifferentiated membership is universal and open to all registered users. Undifferentiated membership has less appeal to users, and efforts can be made to promote these users to higher-level memberships.

2) Members are screened according to the level classification standards. Those who meet the standards can become members of the corresponding level. High-level members can develop into paid members.

3) Paid membership is mainly aimed at high-value members who are willing to pay. Relatively speaking, paid membership has the most obvious effect on user retention and repurchase.

Points system

Points refer to the points that merchants issue as rewards to users for each consumption behavior. Points can be exchanged for the merchant's products or services, thereby attracting customers to make repeated purchases and increasing merchants' profits. Precisely because points can be exchanged for goods or services, points have the characteristics of electronic currency to a certain extent. But in essence, the points system is a marketing of a user growth system, and is also based on user loyalty.

In short: if you want to gain user retention and repeat purchases through points, points will incur certain costs.

(1) Points pricing

First, preset a price, and then calculate (estimate) the minimum and maximum points that can be obtained during the user's life cycle based on this preset price, taking into account the product form, point issuance volume, and point consumption channels. Continuously adjust this value and ultimately come up with a price that keeps costs within a controllable range.

(2) Points recovery

It is meaningless to issue points without determining the recycling channels. In actual operations, the points recycling channels should be planned first, and then the issuance amount and method should be determined.

The following are some common ways to redeem points:

1) Points redemption: exchanging points for electronic currency to redeem products or services is the most common and direct way to consume points;

2) Points deduction: when users support, points can be deducted from a certain amount, which is also a way to directly convert points into electronic currency;

3) Points games, including roulette, points lottery, gashapon machine, Happy Farm and other points games;

4) Product premium. For example, if you need to buy an iPad for 3,300 yuan, you can get 3,000 yuan + 300 points with points. If 1 point = 1 cent before, then 1 point = 1 yuan here, which means that the points are premiumized to a certain extent to achieve the marketing purpose.

5) Regular clearing of points can urge users to actively consume their points, and can also avoid the hidden dangers caused by users accumulating huge amounts of points over the years.

(3) Points issuance

The cost of each point issued should be calculated in advance. Points without cost are meaningless to users (no incentive effect).

Before points are issued, the calculated cost must be reserved. If the amount of points issued exceeds the reserved amount, the points will become inflated. Therefore, you must calculate the cost of points before issuing them. For example, you can take out 1%-2% of the profit as the cost of issuing points.

Common ways to distribute points are as follows:

1) Shopping points. Users are rewarded with points when they shop. The points they earn become silent rewards for users. They can only be activated if they continue to make corresponding purchases, ultimately achieving the goal of promoting user repurchases.

2) Task points. Encourage users to complete things that are valuable to the platform's subsequent communication with other users, such as adding personal gender information, personal age, etc. Obtaining this kind of information can improve the accuracy of personalized recommendations, but also increase the cost of users' silence on the platform.

3) Reward points. When users achieve a certain event, they will be rewarded with points. For example, when browsing products, adding to shopping carts, etc., reward points exist more in the form of point tasks.

Points to note:

In actual use, it is very likely that the points that users spend a long time to acquire will be used up in a subsequent transaction. Therefore, for users, in addition to providing sufficient consumption channels, it is also necessary to make obtaining points interesting and valuable so that users are willing to actively obtain them again after consuming them.

1. The core of e-commerce user growth

Satisfy users' sense of achievement, enjoy privileges and gain benefits from the platform.

For users, a good user growth system can enhance their initiative and activity, and users can also gain a sense of accomplishment, privileges and other benefits from the system. For example, Tmall points can be used to offset next consumption, and QQ super membership allows users to enjoy a different and distinguished experience.

For the platform, the user growth system can help understand how users use the product, which helps to understand the composition of users and can serve as the basis for subsequent refined operations. At the same time, the growth system is conducive to improving user stickiness and loyalty, and promoting user transactions, attracting new users and other behaviors.

In general, the user growth system will be a win-win result for both users and platforms. The user growth system needs to ensure that users receive corresponding satisfaction during their growth process. A user growth system without value is a burden to users, and the final effect is often counterproductive.

Core indicators of e-commerce user growth

To build a user growth system, you need to first establish goals in the growth process, which means key data indicators. Retention and repurchase are definitely one of the key indicators in the growth process. In addition, GMV, average order value, number of orders, conversion rate, etc. can also be considered.

User Segmentation

Different groups of users have different demands, and only by satisfying them differently can we achieve better retention and repeat purchases. Therefore, user segmentation is the foundation for building a user growth system and an indispensable prerequisite for conducting refined operations on users.

Identifying and classifying users with the same attributes to achieve differentiation and make more accurate secondary sales is called user segmentation. Common user attributes are:

• Behavioral characteristics: certain behaviors performed by users, such as activity, number of purchases, usage time, opening frequency, certain specific operations, etc.
• Identity characteristics: based on the user's own attributes, such as gender, region, education level, city dimension, marital status, etc.
• Channel attributes: attributes determined based on user sources. Common channels include Baidu, local marketing, and targeted advertising.
• Sensitivity attribute: an attribute judgment based on user psychology, such as price sensitivity, service sensitivity, etc.

Data collection: Collecting data is the first step in user segmentation. First, determine the type of users to be segmented, then determine the data that needs to be obtained, and do the corresponding data collection work in advance, reward the user behavior data monitoring mechanism, and label the users by type.

Cluster analysis

By classifying different groups of people and then observing the performance of the same event or the same indicator on different groups of people, we can infer and locate the factors that have a significant impact on the event/indicator, achieve a more accurate positioning of user needs, and formulate targeted operation strategies and marketing plans. User segmentation can help companies better understand users and analyze the attributes and behavioral characteristics of different groups. It can also help operations staff better discover the causes behind product problems and find ways to effectively improve and optimize the products.

User Model

User model can also be called "character model", "user portrait" or "user role". Therefore, the essence of user model is user grouping.

The user model mentioned in Alan Cooper's The Essence of Interaction Design is a systematic method for studying users. It is an important tool for product managers and interaction designers to understand user goals and needs, and for development teams and related personnel to communicate and avoid design traps.

The purpose of the user model is to obtain user needs through user research and ultimately serve the product.

(1) User Funnel Model

The funnel model is based on a set of process-based data analysis of users. By observing the conversion and loss at each step in the process, it truly reflects the user behavior status and the user conversion status at each stage from the beginning to the end.

In specific work, we need to look at the overall user funnel, and we can also look at the funnel of certain specific groups of users to identify the reasons for churn and non-repurchase.

For example: the add-to-cart funnel for first-time users and the add-to-cart funnel for multiple-order users. Funnel for non-member users to purchase additional items, funnel for member users to purchase additional items.

Common funnels are as follows:

• User payment funnel;
• Users open membership funnel;
• User add-to-purchase funnel;
• User registration funnel, etc.

In actual operation, you can design relevant funnel models based on your own business.

(2) User life cycle model

The user life cycle model is also a type of user segmentation. It classifies users according to their different life cycles. In actual operation, corresponding key data is established according to the corresponding life cycle, and key nodes (growth period and dormant period) that affect retention and repurchase rate are focused on.

【Growth Stage】The key data indicators in the growth stage are retention and repurchase data. In particular, the process of users advancing from the first order to multiple orders promotes users to become high-repeat purchasing users.

[Mature stage] Users in the mature stage already have a certain understanding of the platform.

[Dormant period] Wake up dormant users through contact channels, give them certain discounts, and prevent them from entering the churn period.

(3) RFM model

In customer segmentation, RFM is a very classic model for hierarchical management of users. The RFM model extracts three characteristic dimensions from the user's business data: the time of the last consumption (Recency), consumption frequency (Frequency), and consumption amount (Monetary).

Through these three dimensions, users can be effectively segmented to distinguish low-value customers from high-value customers, and different personalized services can be provided to different customer groups. Limited resources can be reasonably allocated to customers of different values ​​to maximize benefits.

(4) User Value Model

The difference between the user value model and RFM is that RFM focuses more on high-value users, while the user value model differentiates users based on the "value" standard.

User value will increase as the user grows and depreciate as the user declines. The essence of improving retention and repeat purchases is actually to increase user value.

(5) Establishing a user growth model

The user growth system is built on the basis of the user data model. It forms a complete set of operational mechanisms to drive user growth by finding the key paths and core driving forces of user growth, building incentive channels for user growth and access channels to connect user behaviors.

A good user growth system is centered on products/services, which can help users solve more needs and allow users to achieve self-satisfaction from the products. User growth refers to the different stages that users actually reach in the process of using a product, as well as the different stages that the platform hopes users will reach.

The user growth system generally quantifies user behavior in the form of points, experience points, growth points, levels, etc., and uses this as a standard for judging user activity, loyalty, and contribution. Regulate user behavior through material or spiritual incentives, promote user growth, and thus achieve the goal of enterprise/product growth.

1. User Reach

When users are in the growth stage, their understanding of the product is vague, so it is necessary to strengthen their memory through active contact with them.

Reach Channel

From the perspective of the reach scenario, it can be divided into internal reach and external reach:

(1) External Reach

• PUSH: marketing information, logistics information, customer service information, order information, points information, etc.
• SMS: marketing information, logistics information, order status changes, etc.
• Email (email is still one of the most effective channels to reach European and American users): marketing information, customer service information, order information, etc.
• WeChat groups, WeChat public accounts, and Facebook business accounts.
• Advertising: targeted advertising, outdoor advertising, and media advertising.

(2) Internal Reach

• Internal messages: points information, logistics information, marketing notifications, system announcements, order status changes, etc.
• In-app pop-up boxes: marketing information, promotions, etc.

[Reaching the target audience] Before reaching the target audience, first select the target audience, then choose the appropriate time and channel to reach the target audience. Different groups of people will experience different results when faced with different reach channels, and similarly, different groups of people will experience different results when faced with different marketing content.

[Reaching Channels] Selecting the right reaching channels is the first step to reaching users. Selecting the most appropriate reaching channels according to different scenarios or different groups of people can achieve the best results.

【Reach Time】What time point has the highest reach efficiency? Generally speaking, the time periods with the highest SMS reach are 8-9 a.m. on the way to work, 1-2 p.m. during the lunch break, and 9-10 p.m. at night. In the actual process of calculating the best channel reach time, you can try to conduct a large number of tests to calculate the best reach time for users.

[Effect Tracking] It is necessary to track the reach effect. It is not only a prerequisite for subsequent optimization of reach events, but also required when calculating reach costs. Send volume, click rate (open rate), landing page, number of pages viewed, browsing time, conversion rate, conversion amount, etc.

[Frequency of Reach] Products are also real-life audiences. Putting yourself in their shoes, many people are afraid of disturbing users when reaching out to them. But if they are not reached, users will leave faster. After all, there are many choices and people forget things quickly. Reducing interruptions and reaching users in all directions are not in conflict. You just need to find the best balance through high-frequency experiments.

Notes on reaching out

(1) The most worrying thing about adding new touch events is that they will disturb users too much, so people are very cautious when they first start to use them. However, in actual operations, whether each event should be performed and the frequency of touch can all be determined through experiments to get the best effect. So don’t worry at the beginning, do more experiments and let the data speak for itself.

(2) The best reach time is different in each region. There is a 2-hour time difference between Beijing and Urumqi. If the user distribution area is wide enough, especially if there are international users, you must pay attention to sending during the non-resting time of the receiving users.

(3) Generally speaking, the system trigger time benefit is better than the autonomous trigger event.

(4) The data from different channels are connected. For example, if channel A has already reached the user, then channel B may no longer be necessary.

3. Summary

Retention and repurchase are the future of a product. If users do not retain and repurchase, and leave at any time and are easily lost, it will be difficult to create value for the product or platform. It is because of the value of this platform and the services it provides that your product will be able to create the relevant value and profits it deserves.

Related reading:

1. User operation: new funnel model for conversion analysis!

2. User operation: How to use B-side operation thinking to increase user growth?

3. Product operation: How to use data analysis to drive product user growth?

4. APP user growth: One model solves 90% of growth problems!

5.How to increase users? Take Pinduoduo and Xiaohongshu as examples

6. Triggering user growth: Is user operation just about attracting new users?

7. User operation: What else can you do to attract new users without fission users?

8. User operation: how can financial products awaken dormant users?

Author: Jiang Lailai

Source: Jiang Lailai