The key to growth: How to improve user retention?

Retention , as the name suggests, means that users stay in the product, continue to be active, and effectively contribute to DAU/MAU. In other words, DAU/MAU are the result of new users and their retention. Therefore, after acquiring users, the value generated by users in the app is often related to the depth and length of use. The accumulated value generated by users often increases positively with the use time until the end of the life cycle.

Therefore, it is crucial to effectively extend the user's life span and increase the frequency of activity. Retention analysis can be used to analyze user engagement/activity within the app. Products in various industries also use retention indicators to measure whether their products are in a healthy state.

• New device retention: how many newly added devices will be active in the future? For those with an account system, "new user retention " is also often used (new users: newly registered users);
• New device retention on day xx: the percentage of newly added devices that were active on day xxx;
• New device retention within xx days: the percentage of newly added devices that have been active within xxx days.

The retention used in this article refers to the retention of newly added devices.

First look at the retention curve

After getting the user:

• Early stage: Users quickly decide whether they will continue to use the product based on their experience and choices of the product;
• Mid-term: Whether the product can continue to attract users; products with core value will attract some users to stay and enter a stable period of long-term use.

The figure below is just sample data. The volatility period of most products will even continue until 14 days later.

Improve user retention

1. Retention performance of each channel

There are often significant differences in user retention across different channels, as shown in the following figure: OPPO ’s retention performance is better than other channels. Based on this difference, we can check whether OPPO's user base is special when it is acquired, what the approximate proportion of purchase volume is, etc. When OPPO is confirmed to be a high-quality channel, we can then check whether we can reach an expansion cooperation from the product or company perspective. The purpose is to increase the user base of high-quality channels and thus improve overall retention.

2. Perform retention analysis based on user portraits

Retention analysis is conducted based on user portrait attributes, such as age, gender, geographical distribution, education level, etc. Kuaishou, for example, expanded its user base in third- and fourth-tier cities, created a high-retention and high-growth user barrier, and dominated the video user group until Douyin broke into the market.

If users aged 18-26 show better retention in your product, you can use the results of "whether they are in school or working" to more carefully find "high-quality user groups that match the product and like to be active in the product."

After confirming this group (for example, the student group aged 18-26), you can plan product strategies based on the characteristics of this group, such as "young, campus, immature, and energetic". For example: NetEase Cloud Music purchased more "exclusive music" to attract this group, and Tencent Video created "Creation 101" with the aim of "attracting this group" (based on this group).

3. Analyze the first-day behavior (function usage)

(1) Define users from a functional perspective

From the perspective of user usage behavior (function), compare the retention performance of different groups. For example, the retention rate of new devices that play "2D" videos is significantly better than that of new devices that play "scenery" videos.

(2) Functional retention of new equipment

Functional retention is a commonly used indicator to measure the viscosity of various functions within a product, and improve user retention through high-viscosity functions. For example, behaviors with the same entry and the same operation cost: in a photo community, (1) browsing photos on the discovery page; (2) browsing photos on the friend feed . When the function retention of browsing photos on the discovery page is much higher than that of browsing photos on the friend feed, we must pay attention to the usage trends of the user group, because at this time, the information users give us is that "browsing photos on the discovery page is more attractive to users."

By comparing the retention differences of various functions within the product, we can find the functional points that are sufficient to improve user retention, and then amplify the impact of this function in the product.

As shown in the following figure (the x-axis is function retention, the y-axis is retention standard deviation, and the area is user group): Function 8 has high function retention (among the group that uses this function, the proportion of using this function again is high), low retention standard deviation (stable retention performance), and a relatively large user group. It is an ideal "high-viscosity function", and expanding the scope of influence of this function within the product will be our next step.

(3) Digging deep into the first-day experience process conversion

The user groups with low retention rates can improve the conversion process of each link during their first-day experience and improve the behavioral bottleneck of the product. From experience, the retention performance of users with "deep behavior" is better than that of users with "shallow behavior". Therefore, improving the conversion process of behavior will also improve user retention.

In addition, the above analysis is displayed in units of "days", which is more suitable for entertainment apps such as information and videos. If you are a product that is used monthly, such as: Aunt Flo , you need to pay more attention to long-term retention.

Source: cecil