APP operation and promotion: How to keep users and create value?

Through case studies, we explain in detail how to keep users and generate value. The full text is divided into 5 parts:

1. The essence of retention
2. Improving the foundation for retention
3. Prevent user churn
4. Promote user activity and generate value
5. Recalling users after churn

The essence of retention

The essence of retention is to extend the user life cycle ?

This statement is correct, but not comprehensive, because if a user is a "troublemaker", for example, someone who only leaves malicious negative reviews every day and adds no value to the product, then this type of user is not the target user . Therefore, the essence of retention is to retain users and bring greater commercial value to the product, rather than just extending its life cycle.

Improving the foundation for retention

Taking the operation of the catering industry as an example, whether it is Haidilao or other restaurants, the first thing to solve is how to attract more users to dine? The prerequisite must be that the food must not be unpalatable, which means that the product must meet a real need of the user. If the product itself is a false need, then no matter how the product is operated, the retention rate will not be high.

Prevent user churn

Next, I will talk about how to prevent user churn through some specific cases.

First of all, users can be divided into retained users and lost users .

Secondly, each user has two dimensions of attributes:

• User attributes: age, occupation, gender, etc.
• User behavior : browsing, viewing, ordering, paying, etc.

Combine user attributes and user behaviors to conduct refined operations on users, identify the behavioral differences between retained users and lost users, and find out the reasons for loss? Is it gender? Is it the browser the user is using? Or are there certain features that are triggered by retained users but not by churned users? After finding the differences through user portraits , corresponding revisions and guidance are made based on the differences.

Case 1

When a certain start-up vertical e-commerce company was first launched, it spared no effort in promotion and provided large user subsidies. However, it discovered that after the subsidies were cancelled, user churn was very serious. Therefore, it was necessary to analyze user behavior data to find the key product points to improve user retention .

Category navigation entrance advertising space activity entrance (for illustration only)

First, through user grouping , filter out users who have made purchases, and then divide these users into two groups: one is users who have purchased and retained, and the other is users who have purchased but churned. Based on user behaviors, such as successful payment, the data can be further refined and divided into two groups for comparison.

After separating the two groups, first check the user portraits to find the differences. Through observation, it is found that the number of times the two groups of users trigger the category navigation is different. At this time, the two user groups can be compared to determine the activity of triggering the category navigation events.

As shown in the figure, the blue columns are retained users, and the yellow columns are lost users. The blue column is much higher than the yellow column. At this point, we can put forward a hypothesis: users who have triggered the category navigation have a higher retention rate . Then, we need to verify the hypothesis next.

During verification, users were also divided into two groups to see if there were any differences in the category navigation events.

• The first group includes users who have triggered the category navigation most times within three days;
• The second group includes users who have triggered category navigation less frequently within three days.

Through the funnel function , we found that the blue columns represent users who trigger the category navigation more times, and the yellow columns represent users who trigger it less often. Obviously, the conversion rate of blue users is much higher than that of yellow users .

Looking at retention, the blue curve represents users with more triggering times, while the yellow curve represents users with fewer triggering times. The blue users are nearly 20% more than the yellow users. This verifies the previous hypothesis: users who have triggered category navigation do have a higher retention rate than users who have not triggered category navigation.

From a product perspective, the role of category navigation itself is a form of content organization. It is a way to improve the search efficiency of individual users, bring a more comfortable experience to users, and allow users to find their favorite products faster and complete relative conversions. Therefore, the company revised its products.

Promote user activity and generate value

High frequency function drives low frequency function

The core functions of a product do exist, but the product itself is a low-frequency and infrequently used product. For example, a travel product is a product that meets the real needs of users but has a relatively low frequency and is not often used (users travel an average of 2-3 times a year). At this time, a high-frequency auxiliary function can be added to serve as the traffic controller.

For example, if a community is added to a product, and users who do not have travel needs can also log in to the product to view travel guides and earn points by checking in and other actions, then this community is an auxiliary function, not a core function.

By combining the two parts, we get a new product that users can use frequently and meets their core needs.

Example:

Vanke APP—live here. The initial functions of this APP are: personal butler, visitor invitation, mobile phone door opening, and mail order. Although these functions are very practical, they are not necessary for users.

For example, the door opening function with a mobile phone, in fact most owners use room cards. Vanke added a community function to guide users to complain, for example, whose dog poop was not cleaned up in time; where there are cigarette butts; or organize owners to participate in fun activities. In this way, through the participation of owners, the activity level is increased, and ultimately the retention rate of this low-frequency APP is improved.

Increase the cost of leaving the user

In love, the one who gives more is often more unable to extricate himself. In fact, the same is true for products. Letting users give more will make them reluctant to leave.

Little Yellow Car

I joined an official WeChat group of "Ofobi", which provides some guidance to win the sympathy of users. For example, the maintenance technicians of Ofobi have a hard time. Users in this group will also make comments like "Poor Ofobi, I woke him up lying in the grass". Some users even volunteered to stay up all night to help the operation and maintenance personnel collect the bikes. I believe that when these users need a bike, they will definitely choose ofo first. This is an example of how users will become emotionally attached to a product and will not easily leave it.

Mobike

Initially, Mobike supported recharges starting from 1 yuan, but later adjusted it to a minimum recharge of 10 yuan (the refund process was imperfect and criticized by many people). By recharging in advance, users are bound and cannot leave the product.

Zhihu

If a user has answered 149 questions, and all of them are of relatively high quality, if there is a new platform at this time, he is unlikely to leave Zhihu, because the cost of changing platforms is also very high.

Xianyu

For example, I have uploaded a lot of products that I need to sell on Xianyu, but now "Zhuan Zhuan" has come again. If I want to switch to "Zhuan Zhuan", I need to re-upload all the product information previously on Xianyu. This workload is huge. I have already paid a lot of costs on Xianyu, so increasing the user's leaving cost can promote user activity in the product.

Build a user incentive system

• Step 1: Clarify user roles. If it is a content product, there are content producers and content consumers. What is their position in the product? If it is an educational product with teachers and students, then the means, methods and measures of motivation will be different.
• Step 2: What are the behaviors that are motivated? Some of the user behaviors in the product are valuable, for example, purchasing goods is valuable for e-commerce platforms; some behaviors may be harmful, such as making negative comments in social products, which may lead to the product being blocked.
• Step 3: How to motivate? Just satisfy a weakness of human nature.

for example:

1. Weakness 1: Spiritual satisfaction, for example, the level of Taobao buyers;
2. Weakness 2: Material satisfaction, for example, Ele.me points mall;
3. Weakness three: being different from others in terms of product functions (being unique). For example, in Baidu Tieba , some posts only allow users above level 4 to post, which will give users a sense of vanity and superiority.

• Step 4: Choose a suitable method.

First, clarify what users want and combine it with what the product wants, that is, find the balance between product needs and user needs, and evaluate the user behavior required by the product.

Secondly, take advantage of the herd mentality to make users feel that many people are using it, which stimulates users. For example, the “Success Stories” section in the Bohe APP stimulates users to join in through stories of successful weight loss.

Increase the number of available scenarios

Two years ago, Didi Dache was renamed Didi Chuxing, and the name was not changed casually. When Didi first started, it was called Didi Dache because this name could make a deeper impression. Before, users only thought of Didi when they took a taxi. Later, as Didi’s popularity increased, it was renamed Didi Chuxing. At this time, users would think of Didi whenever they wanted to travel, which expanded the user’s usage scenarios compared to before.

The same example, for example, Wanglaoji , was previously called "Cure getting a sore throat, drink Wanglaoji", and later it was changed to "Avoid getting a sore throat, drink Wanglaoji". If you have a sore throat, it means you can only drink Wanglaoji if you have a sore throat. But if the name is changed to "Drink Wanglaoji if you are afraid of having a sore throat", then at this time, as long as users intend to prevent having a sore throat, they will choose Wanglaoji.

Recalling users after churn

If all the previous measures have been taken but users are still lost, user recall is necessary at this time. However, user recall has costs, and it is impossible to recall all users because some users are inflating their numbers and some are junk users. These users should be abandoned and the real users should be recalled first.

If the initial product does not meet user expectations, it is likely that users will be lost. After the loss, you need to find a way to recall these users.

Similarly, use the user grouping function and select a filtering condition: users who have not logged in in the last 30 days are defined as lost users, and real-name users who have made payments three times in the last 30 days are defined as high-quality users. This user group is named high-value lost users, and now this part of the user group needs to be recalled.

Common methods of recall include email, product push, text messages, and WeChat service accounts. These methods have obvious advantages and disadvantages. For example, although the SMS delivery rate is high, the cost is also high, and the current SMS usage rate is not high; the push of service accounts has a high delivery rate, high opening rate and low cost. But the disadvantage is that it is very offensive to users and users may not like it. Therefore, it is necessary to choose the recall method based on the characteristics of the product.

Mobile application product promotion service: APP promotion service Qinggua Media advertising

This article was compiled and published by @朱葛io (Qinggua Media). Please indicate the author information and source when reprinting! Site Map