What is a good retention rate? 3 cases to analyze the core direction of user retention

The identification standard of retained users recognized by the Internet industry: regardless of the channel through which the user enters and uses the product, if the user continues to use the product after a period of time, the user is considered a "retained user".

How to calculate user retention rate : Retention rate = number of new users retained / number of new users on the day * 100%

What is a good retention rate? The retention rate of each product is different at each stage. For some products, a 30% retention rate is enough to cover the customer acquisition expenses, while for others, a 70% retention rate is required to cover the customer acquisition costs. The judgment criteria should be combined with the cost-benefit of the product to find the profit balance point, and strive to achieve or even exceed it.

The retention rate data of APP products is not convenient to use as a case study. Here, I will use the data of my own WeChat public account in the past week as an example to introduce the " user retention " thing (the data is real)

1. Statistics

The APP side usually uses data collection to count the number of new users and their retention. Of course, if you want to develop a WeChat public account side, you can also do it. I am really good at it. The amount of data is so small that it is naturally counted manually.

Although it is a manual statistics, it also requires skills and it is definitely not just counting one by one.

The statistics of the number of new users on that day are available in the official account management background and can be used directly.

(The picture was taken on October 19)

The user retention statistics of new users added on a certain day within a week require the use of the tag function. Tags are used to label the users you follow. To count the user retention statistics of new users added on a certain day within a week, you need to use the date as the tag and mark all the new users at that time with the date of that day. As shown in the figure, for example, 4 new users were added on 2017.10.11, so I marked these 4 users as 2017.10.11 to represent that the users were added on that day. The public account background can be used for batch operations by making label markings.

When the new users added on 2017.10.11 are lost , the number of users labeled 2017.10.11 will decrease. The current value is the remaining number of new users added on 2017.10.11 on this day. The data is counted in Excel one by one, and the above formula is used to calculate the retention rate of new users added on 2017.10.11. We generally focus on the 1-day, 3-day, and 7-day retention rates. Based on the retention rates, combined with recent operations and user behavior data, we identify the correlations and key factors that can encourage users to stay. We then make them into operational documents as the focus of subsequent operational strategies.

2. Data Analysis

As shown in the figure, the number of new users of the official account is increased every day.

The new user data is consistent with what happened on my official account in the past week. On October 12, Zhangji Grocery Store shared "How to build a personal brand? 》 brought a lot of traffic to the account with thousands of followers, and directly generated attention conversion . Here I would like to express my gratitude to Brother Liang (the chief writer of the grocery store). In the next few days, the number of new users gradually returned to normal, which also verified the effectiveness of large accounts in bringing traffic. I will continue to polish the articles and welcome friends from all walks of life to continue to support me.

The figure shows the retention rate of new users added every day within a week.

The user retention data is very ideal. Normally I update one article a week. When there was an increase on 10.12, I deliberately updated 3 articles in one week. In addition, the welcome package for users includes financial, medical and other data reports as of October 2017, 50G of Internet course materials , and a large number of high-quality Internet-related books in PDF version. These show that at least I am a person who can obtain resources in a timely manner, which is helpful for users to stay.

2. Reasons for users to stay

User retention in the Internet industry is the same as retaining customers in the physical industry. The essence of keeping users to stay and use your products and keeping customers to visit restaurants next time is the same. It is even the same as the essence of getting along with people. In layman's terms, it means "matching personalities", "suiting your appetite", and "giving surprises". To use three common words: like, identify, and expect.

I feel very lucky that the users of the official account commented on my account: "I like your official account very much. It is warm and informative ."

When the first user said this, I felt flattered; when the second, third... and tenth users made the same comment, I began to feel a little confident about the official account; by the 20th positive comment and no negative comments so far, I completely began to believe that "my official account is really both warm and informative." What a high evaluation! For a personal account like mine, praising the official account is a direct compliment to me. Thank you all very much for your high comments. (It is not convenient to take screenshots in private chats)

Through the feedback from users, I found that they really "like, identify with, and look forward to" my official account. They like the writing style and articles of the official account, identify with the views and skills of the articles, and look forward to more high-quality articles. It doesn’t matter whether it is a hit or not, as long as the content continues to maintain its style and quality, they believe, as I do, that I can slowly make the official account better and better. I really believe in this and I am very grateful for their continued companionship and encouragement.

Case 1: Xiaomi and Mi fans’ mutual admiration

Liking a product is the highest evaluation, just like painting a picture of a star you like. Users who like your product must be silently helping the product grow. When your product has a bug, they will report it to you immediately and will not uninstall it. When your product wins an award or is commercialized, they will be happy for you and applaud you from the bottom of their hearts.

According to ZDC's research report on China's smartphone market in July 2010 in August 2010, Nokia , HTC , Samsung , Apple and BlackBerry accounted for 90% of the market in China. After checking relevant reports on mobile phone sales in Zhongguancun, we learned that the most classic models at that time were all very expensive. Among them, the 16GB Apple iPhone 4 was priced at around 5,200 yuan , the Meizu M9 (8GB) was around 2,500 yuan, the Samsung I9008 was 4,500 yuan , the Motorola Defy was 2,300 yuan, the Nokia N8 was priced at 3,800 yuan, the Coolpad N930sh was priced at 6,000 yuan, and the HTC 7 Trophy was priced at 3,900 yuan.

The above background can reflect that in 2010, mobile phone users could not buy a good mobile phone for 1,000 yuan. Against this background, Xiaomi 1, a high-quality smartphone specially designed for enthusiast-level mobile phone control, was launched and priced at 1,999 yuan. The differentiated competition of "a thousand-yuan phone can also be extremely powerful" won the favor of users. Buying a high-quality smartphone was a very cool and face-saving thing at that time . Xiaomi gave face to users and Mi fans supported Xiaomi.

In its subsequent growth, Xiaomi has always followed the vision of " allowing everyone to enjoy the fun of technology ." It created a model of developing mobile phone operating systems using the Internet model and in which enthusiasts participated in the development and improvement. "Born for enthusiasts" is just a slogan to passers-by, but it is a promise to Mi fans.

It is the promise made by the engineer “If you have any questions, just come to me” when Mi fans praise the engineer and say “Long live a certain engineer”; it is the promise made by Xiaomi’s product, technology, marketing and operation colleagues to put Mi fans first after Mi fans have built the world of Xiaomi through their ultimate spiritual power.

"When you truly trust your users, they will also trust you." The first 100 fans on the Xiaomi forum were also the igniters of the MIUI operating system. Later, Xiaomi released a special micro -film to thank those 100 die-hard fans. It was called "Sponsors of 100 Dreams". Their names were projected onto the big screen one by one to express gratitude to them. At that moment, many of them burst into tears. Xiaomi really believes that Mi fans can help them independently develop an operating system that belongs to Xiaomi and Mi fans, and Mi fans also trust Xiaomi. This is also a recognition of the Xiaomi team.

To this day, in the Xiaomi forum, you can still see Mi fans constantly raising questions about Xiaomi products. Although Xiaomi was criticized by Pinwan for releasing new products this year, Mi fans still support Xiaomi's products. They expect Xiaomi to develop more extreme hardware products. This expectation is the result of long-term liking and recognition of the products.

So, when we look back at our products, have we ever seriously thought about why users follow us? "The product tools are easy to use", "The product content is interesting and profound", "The team behind the product is interesting", "Everyone is using it". If our products are not as "easy to use" as Shimo Documents, not as "profound" as everyone is a product manager , and not as "popular" as WeChat, but have the "cumbersome" user rights verification of Toutiao 's mobile phone number every time, and Soul's "too niche", then the retention rate should be difficult to improve, unless one day the general public's preferences begin to tend to "explore complex products when bored, and find the most niche products to prove their uniqueness."

Case 2: Mi Meng ’s chicken soup is “toxic”, but users are happy with it

The evaluation of Mi Meng circulated online: " Rookie interns regard her as a professional and intimate sister, friends regard her as a gossip encyclopedia , and netizens regard her as an idle nouveau riche."

Mimi Meng is different from Fan Yusu, who accidentally went viral on WeChat Moments because of her article "I am Fan Yusu". As her fellow villager, she is very happy that Fan Yusu's articles are liked by everyone. There is also such a person in her hometown of Xiangyang, Hubei who persists in the ordinary world, experiences the hardships of life firsthand and writes simple and good articles. Fan Yusu has not received any writing training. Of course, after that article went viral and various media interviewed her, there was no news about her on the Internet. Mimi Meng graduated with a master's degree in Chinese from Shandong University. She worked at Southern Metropolis Daily for 12 years as chief editor and is well versed in all aspects of writing.

The comments under every article of Mimi Meng are "emotional identifiers" . She is very good at writing articles about things happening around her, with a friendly and natural beginning and a powerful ending. She is a pure intellectual elite who is not influenced by the secular world or personality evaluations. In fact, she has separated "self" from "my articles". Like all Internet elites, she has a systematic writing methodology, user analysis methodology, and marketing methodology. So don't think that you can succeed casually . Here is a warning to those who just write articles and hope that others will like to forward them: "If you are not influential in real life and have your own traffic, not many people will like to read your casual daily complaints, unless they like you. 』

Mimi Meng writes about various "problems" in people's lives. Not only does she dare to speak out about the problems, she can also develop relevant courses to help users solve them. Users not only agree that she has expressed their own thoughts, but also expect her to provide a systematic solution . Mimi Meng and her users recognize, accompany, like and look forward to the future together in a world that we think is "toxic."

Every self-media product is spiritual food for users. It can be in-depth information like "LinkedIn" or spiritual reading like " Ten Points Reading ". In short, it is necessary to pay attention to the next tag that users like in their minds.

Case 3: The deep bond between NetEase Cloud Music and its users

I once wrote a review of NetEase Cloud Music:

"In addition to the sentimental advertising done well, NetEase Cloud Music's product UI layout of red on top and black on the bottom is both enthusiastic and affectionate (increasing the visual experience when selecting songs), popular UGC comments from top users share moving stories (setting a sentimental tone for the product, increasing the opening rate and length of stay), promotional videos shot in cooperation with brand sponsors (novel without affecting the visual experience, occasionally stunning and touching, giving users a sense of freshness), celebrity anchor FM (even if you can't stay up all night, there are music and stories to accompany you and understand you, enhancing users' stickiness and awareness of the product, expanding from a music product to a high-quality "companion product" with attitude and sentiment that integrates music, FM, MV videos, and exclusive interviews), the birthplace of niche singers, not so high entry qualifications, some users brought by social networking (a complete music library does not mean that the listener needs it, listening to songs always impresses oneself, and NetEase Cloud's songs are often just right in terms of time, place, and people), a pioneer in playlist classification (users create playlists spontaneously, as long as they are popular enough, they can be on the homepage), and the most important thing is that I like it and it is easy to use. I can accept it even if I download and import it from somewhere else and listen to it again."

At the end of September, I received a letter from Q-misky, the editor-in-chief of the product topic of Jianshu . The general message was that the article about NetEase Cloud ranked first in the number of comments in the product topic. I went back to read the comments of Jianshu readers and picked out a few representative ones for you:

• @Asirnmo: Apart from the money Tencent has to spend on copyrights, I can’t find the feeling of listening to music.
• @小小同学儿: I use NetEase because I like someone who uses NetEase Music, because I want to get to know him, and also NetEase has a radio station with the channels I like.
• @随风话雪: There are problems, there are emotions, and commercialization is not cold.

The users' feelings are real, and the feeling of listening to music is closely related to the product's vision and tone. The commercialization of NetEase Cloud Music has not affected the overall functional layout of the product, nor has it been forcibly implanted. It has transitioned from a membership system to NetEase Yanxuan , and has collaborated with celebrity artists in concerts and albums, and has cooperated deeply with brand sponsors on music. It has always been a music-centric commercial model in which profit and user experience coexist.

The most direct factors that make users feel emotionally resonate are that NetEase Cloud products are "easy to use" and "interesting". Users stay to make some efforts for their original emotional sustenance and look forward to the return of NetEase Cloud after a turnaround.

3. How to keep more users

Whether our retention rate is good or bad now, we can do better.

1. Understand the process of users coming and going

The coming and going of users in a product is actually the same as the friendship or distance between people after getting along with each other. At first, they have a good feeling about "you" in a certain place, so they try every means to get to know you. After getting to know you, there are two directions. Those who "match the initial impression" will continue to observe you, and those who "misjudged" will not only leave but also block you. This is also the fundamental reason why some users will ignore you no matter how you call them back.

This process is cyclical for a long time when users initially pay attention to the product. Even if you do something that does not conform to their cognitive feelings, users will leave. At the end of this process, if the user is still there, it means that she is completely devoted to you, just like NetEase Cloud Music users cherish you as much as NetEase Cloud users.

If I have to describe it theoretically, I would create three words : "judgment period", "verification period", and "coexistence period".

Judgment period: When users first enter the product, they will explore the product's functional system under the guidance of a novice or on their own. They may try to use one or more functions. Once they feel that they are not suitable for them, they will quickly judge to "leave", otherwise they will "stay and see".

Verification period: The decision is made during the judgment period. The user thinks the product is good and there are some points that she values, so she stays to continue observing. At this stage, the user will verify whether you meet her initial judgment. Once there are points that do not meet her initial judgment, the user will "leave".

Coexistence period: After a long verification period, users confirm that the product is very suitable for them. At this time, users are not so easy to leave. Even if you make a small mistake, users will think about how to help you recover and will not easily abandon you.

I always believe that the process and standards of "girls' choice of boys" in dating thinking and "users' choice of products" in product operation thinking are similar. Understanding and guessing the user's psychology can simulate the selection criteria between people.

2. Core methodology to retain more users

① Find out what actions will increase the probability of users staying

Take my official account as an example. "Increasing update frequency", "maintaining article quality" and "giving away more information that the general public cannot obtain" are the key points to attract more users. The reasons for these three points have been analyzed in the previous article.

Case: Analysis of user retention behavior of Yidian Zixun

From the data of a large number of users coming and going, we found that most of the people who forwarded the content recommended to users outside the site ( friend circle , Weibo, etc.) are still using the product, and the retention rate is very high; the users who stay on the product page for a very short time and keep jumping between multiple product pages have serious churn, because they will "flip over and over again" when they can't find the content they like. As for how to discover these features, we need to group users according to their behavior, compare "shared >3 articles within 7 days" with "shared <1 article within 7 days", and compare "page dwelling time <3 seconds" with "page dwelling time >10 seconds". In this way, we can show obvious differences in the retention rates of different groups.

Therefore, there are two things to improve the product. The first is to work with the development, product, and data-related departments to optimize the intelligent recommendation algorithm, and strive to recommend content that users are interested in to them; second, to upgrade the content of the authors of the product system, strive to attract more high-quality authors from Zhihu, WeChat public accounts, and Baijiahao , and give them more generous subsidies, so as to improve the content of the product page and increase the time users stay on the page. Use the above two strategies to improve the retention rate of Yidian Zixun.

② Actively extend user demand for products

Especially for tool products, if user demand is only present in a certain period and the frequency of reuse is very low, it will lead to a very large user loss. At this time, I think the user demand should be extended.

Case: My opinion on the way out of driving school

The previous article has analyzed the user retention problem of Jiaxiaoyitong. The main problem is that the "user demand curve is too short ". After users pass the driving test, this product has no usable value for users. So the core of my advice at the time was to "extend the user's demand curve."

Many tool products are relatively isolated. If you want to extend user demand, you have to seek more external product cooperation. In fact, this is to divert external traffic and cannot fundamentally solve the problem of user retention. However, Yidiantong is within the 58 system. 58 has very strong offline car resources and 58 Finance's financial resources. If Yidiantong + 58 Finance + 58 Suyun form a car loan business with passenger and freight drivers as the core, from learning to drive to buying a car and pledging the car to 58 Finance, the drivers can make money through passenger and freight business to repay the car loan. This is a relatively clear direction to extend user demand. As for whether it can be implemented and whether it will be effective, I don't know.

Having written this, I have finished writing about the relevant ideas on user retention rate. I still hope that you can make some comments on the storytelling nature of the article and the comprehensiveness of the knowledge.