The author presses: The core of the user operation, is the active user's scale to raise, the way has two kinds: open source and the throttling, opens is registers the source, the section is drains the flow. There is also one, that is active, let inactive users become active, let active users more active. This article is divided into three parts: "Open Source: How to increase the number of registered users", "throttling: How to retain registered users" and "crazy operation: How to keep users active".
This is a very practical value of the long article, the author of Sina Weibo operations Manager Jin Pu (@ Little Yan). In order to ensure the continuity of the Mobile 2.0 readers to the long article reading experience, we chose to post the full text at once, supplemented by the table of contents jump way to carry on the page display.
Directory
Open Source: How to increase the number of registered users
One, binding registration brings more users
Second, how to do a good job of registration transformation
Throttling: How to retain registered users
Throttling (i): Building a user model
1, how to define user churn
2, how to build the user model
Throttling (ii): How to prevent users from losing?
User Churn Alert
User Churn classification
User anti-loss
How to recall the lost user?
How do I make sure the content of the recall information?
Improve recall effect
Lead after recall
Crazy operation: How to keep users active
Establish active measures according to user registration time
Allow users to increase user activity
Maintain active user Activity
Core User Operations
Summary
Open Source: How to increase the number of registered users
Keywords: Open Source registration transformation
In the open registration channels, improve user registration, all products are spared no effort to grab the entrance, and then improve the conversion rate.
One, binding registration brings more users
After years of development, the Internet product registration channels are more mature, in addition to the product itself registered entrance, there will be and some outbound product cooperation, business promotion, binding registration channels. The first two ways have been very mature, the last way the application is also very broad, for the new or small users of the products is a good way to a big spender.
A binding registration is a user who is registered on another product and is considered a registered user by authorization. This registration method can significantly reduce the cost of obtaining users. For example, the current Third-party application of the login page, tips can be used Sina Weibo, QQ, MSN and other account directly login, is a way of binding authorization. This approach can quickly increase user size.
The so-called "no forever friend, only the eternal interest", even if your product and a group of a product is a direct competitor, your product can and the group's other products have cooperation opportunities. As long as the two sides to find the point of mutual win, cooperation naturally.
Second, how to do a good job of registration transformation
Registration guidance is an important link after user registration, and is the first process of educating users. Registration transformation is a detail job that makes the product manager feel a lot of headaches.
Typically, we need to guide the process by telling users what our products are, what the core values are, what services are available to you, how you should use the product to meet your needs, and so on. The product in the Registration Guide link will be divided into several steps to tell the user the above content. Only by registering the process, the user can reach the user interface of the product, so that the users can be transformed into a real one.
However, the loss of this process is quite serious: from the data statistics, each additional step registration guide, will allow the registration conversion success rate decreased by 10%. Successfully registered users after 5-step registration guide, basically only half of the left. But if the registration guide process does not do, it is difficult to find the right opportunity for users to fully understand the product, then the user because do not know how to use and lost the probability of a significant increase.
Therefore, we need to consider how to balance, on the one hand, consider the full introduction of product features, on the one hand to prevent users from losing. Because the difference between different products is larger, this can not be said that the registration of the boot process a few steps more appropriate. But it is necessary to grasp the following principles:
1, according to the function of the product combined with your product hope that users how to use, to find the highest demand for users, in the registration and guidance process to maximize the display of product core value, to meet the highest demands of users;
2, the function of the secondary function, icing on the cake to wait until the user came in to introduce and then show.
3, through the user's real data, through the adjustment of Copy, function Introduction order, and so on to optimize the registration process, so that more registered users to truly convert to use users.
Throttling: How to retain registered users
Key words: Throttle user Model user demand to prevent loss after the recall
Throttling (i): Building a user model
Open source method is more mature and fixed, the enterprise's industry resources and market resources, business manager negotiation ability and execution ability have how strong, will determine how many products can be available to register the entrance. In contrast, the work of throttling is more diverse and minutiae: why users lose, how to prevent loss, lost how to recall it?
1, how to define user churn
Different products, the definition of the loss of users is not the same: such as micro-bo, happy net, such as SNS website, one months do not use the user is defined as loss, and such as Taobao, Jingdong and other electrical products, due to the specificity of user behavior, may be 6 months of users will not be defined as loss
We should use the product to use the user's real data, build the lost user model for the product, in order to define what user can be counted as "the lost user", what user has the loss trend. Answer these two questions, then how to prevent the user to lose the means to be solved.
2, how to build the user model
According to the previous user's use behavior, calculates the user to use the product life cycle. This cycle will include when users register, after registration in a certain time window to become mature users, proficient in the application of products, and frequently used, until finally become the loss of users, from frequent use to almost no use.
Of course, some users have never reached the skilled application of the time has been lost. If this proportion is large, need to reflect on whether the product has a mishap.
The loss and registration of two data combined to look like a primary school Olympiad water Pool water storage problem, see you store fast or put quickly, you can know whether this swimming pool can be filled with water or shining.
Is your pool water or water?
General user behavior will be a normal distribution of the curve, when users use the cycle is determined, many results will be read from the data. The following figure is an example:
From the curve, we can users from the registration of new users to mature users, is a slow upward process, the same, from the loyal user to the loss is also a slow decline in activity process. The so-called prevention of loss, that is, if the product can be adjusted in time according to user characteristics, start a certain loss of early warning plan, I believe can retain a part of the loss of users.
Once the individual user's usage cycle is determined, the loss trajectory can be depicted for the overall user of the product. Example of a user registered on the same day:
The concept of loss must be prolonged and unused, which needs to be distinguished from the concept of the retained User: The statistical method of the retained user is generally used as the definition of the user retained for the first few weeks, rather than being retained in the previous weeks.
From the chart analysis we can learn:
1, when the curve tends to be smooth, it shows that our natural loss tends to be smooth, that is, if the user is still retained, basically he will be able to retain a long time, more stable, and we can use this point to define our loss. As shown above, we can define the product user if not used for five consecutive weeks, can be judged as loss;
2, natural wastage rate: The above figure for example, the same day registered users 100, of which 20 in the five weeks after the total no use behavior, then the day of the registered user natural loss rate of 20/100=20%;
Throttling (ii): How to prevent users from losing?
User Churn Alert
From the user churn model, we need to analyze the loss of users before the loss of what the same characteristics: if the registration channel is more centralized? What is the sex ratio? is the geographical feature more obvious? Does the age level converge? Are behavioral characteristics more similar? are property characteristics the same? Wait a minute.
For example, the proportion of the lost user to upload avatar is 20%, while the same registration time of the retained users upload avatar ratio of 80%, then the upload avatar is a very important key indicator. (Of course, can also analyze the upload avatar of the user, for the real head of the account.) )
A lot of products think that upload avatar is just a dress up small function, does not involve the user's core function use. Little does not realize that small head has a very obvious butterfly effect. The same SNS community as an example, upload the avatar can in the virtual community to let each other in the first glimpse of the virtual ID behind the preferences, if it is true avatar there is greater probability of real friends to recognize their own, for users faster and better build user relations have an indelible role.
Preliminary judgment: This is an anime enthusiast
In the SNS community, users want to leave less personal information, but want to find more friends. The result is to come up to look for a half-day friend, but not be recognized by friends and ignore, unknown loss. And such users because of leaving too little information, the product mechanism is also difficult to help them find relationships, build circles.
Therefore, for social products, do not give up such a user, each registered user is to give the product a display of the core value of the opportunity, desperately to show the same time, do not forget such as uploading avatar these can let the core value show the small function of a multiplier.
User Churn classification
After finding the key index, we can classify the lost users, find out the different operation strategies and schemes for different classification users, and develop different prevention and control plan.
The key indicator is not only the analysis of the lost users, but also the focus on finding the users who have the tendency to lose. That is, when the user's loss of key indicators began to decline, indicating that the user is a potential loss of users, the need for early access to release the loss of early warning strategy to prevent its total loss.
If the user has been lost, and then come back will be more difficult than the new user registration, new user registration can be because of fresh try, and the old user loss is ungrateful, recall such as the spiritualism, easy.
User anti-loss
From the previous definition and classification of user churn, we focus on the loss of the user's behavior characteristics analysis, designed to identify potential loss of users in advance. So what happens when you find it?
Just now we are looking back at the loss of users, it is better to look at the retention of high-quality user growth path. Why do they play well? Did they have a knack or what kind of doorway they found? We can find the reason for the user to stay according to the data of the high quality users.
After all, not all of the retained users are the same, so we have to the retention of users also need to classify, in the classification method can refer to the retention of the user's registration channels, behavioral characteristics, geographical and so on;
Do you have the same label as me?
For example, register channel A's user registers to like to play the game, then the same channel registers latent loss user may guide to play the game;
For example, some SNS users rarely interact with others, but are active retention users, it is necessary to analyze the reasons for his retention, this is also not like to speak of potential loss of users may have a very good reference meaning.
Regional characteristics can not be small peep, Guangdong area friends and friends in Beijing to use the product of the habit is really a difference, concrete doorway still have product data to speak for themselves.
How to recall the lost user?
The industry to use more recall the loss of the user's means is also relatively single, according to the general use of registration information recall, SMS recall, through the client-side notification of the way recall, and so on.
Although the client's way depends on whether the user chooses to allow push in the installation, but this method generally results in better, because this is the most direct way to find, the window is very annoying, but the effect is wonderful!
Would you order "details" or "Cancel"?
The sending cost of the message is low and the volume is high. However, the turnover rate of the mailbox itself is also very high. Chinese netizens are not very dependent on the use of mailboxes, in addition to work needs, very few users through the mailbox to contact feelings, deal with life affairs. Even many users registered their mailboxes only to register another product's account, and have not logged in since. Therefore, the volume of mail sent is large, but the general open volume is not optimistic, not users do not want to see, but users simply do not see.
The arrival rate of SMS is very high, the disadvantage is high cost, the probability of being treated as spam message is high. At the same time, because users generally have a strong concept of privacy protection of mobile phone number, SMS recall can easily lead to user complaints, so need to use caution.
How do I make sure the content of the recall information?
By means of making the recall message/SMS/push message more effective, you have to make the user feel that it is a gift rather than a nuisance. To change the popular point said: You have to let the product's active push to find the user's G-point, so that users happy! The best is to let users look forward to the arrival of the push!
How to find the user's G-spot? Each product because of different characteristics, different target users, the demand for products is different, so can not be generalized, but I think there are several general principles:
1, such as group buying, electrical products, customer demand for products is convenient, affordable and safe. Therefore, the push should pay attention to meet the needs of users. Push the time is not to tell the user you have today the client changed what new skin, but today what users are interested in discount activities, mail items and so on. Push at the same time, to solve the user has not been able to find such a good discount activities demand;
2, microblogging and other SNS social products, user demand for products are friends, the right to speak, information access. So push is not to tell users what the product fixes bugs, but whether there are new friends to find you, contact you, you are interested in the hot spot again, you want to pay attention to and so on. Although the initiative is to push, but the solution is that the user is not in the net, what happened he wanted this product to get something.
3, entertainment, news and other information products, user demand for products: information access. Push the time except hot, without him. Do your own product experience and information acquisition accuracy, timeliness, you can.
Improve recall effect
Let's start with the most common mail recall to see how to improve the recall:
Mail recall focus can look at three indicators: the success rate of the message, the rate of open mail, mail user regression rate. The first two are well understood, and the third refers to the user-account ratio of users who reuse the product by recalling the message.
There are many factors that affect the first indicator, and one basic principle is whether it complies with the Anti-Spam Alliance specification. In addition, the details to be considered include the appropriateness of the title, whether it is within the range of the sending server, and whether it meets the sending frequency requirements of the receiving party.
The most important factor affecting the second indicator is the title! Title attractive, open on the high, in addition to the title content to be attractive, the title of the details of the processing is very noteworthy. Considering the different habits of the mail users (through the Web/using Foxmail/outlook), the same message title may be seen by the number of words often wide gap. Therefore, in the title, the more the position of the former is more precious. And news headlines, the key word before, is a need to pay attention to a problem, do not waste the former 6~10 words of the precious layout!
You order him and you win!
Then look at the return rate of mail users. The promotion of the first two indicators has a fundamental effect on user regression, while the quality of the message content plays a decisive role in promoting it. How to quickly show the core value of the product in the recall content, make the user feel cared is very important. Here are a few things to do: (1) analysis of the loss of users of the core demand for products, (2) analysis of the loss of users ' interests and hobbies; (3) Analyze the behavior characteristics of the lost users; the above preparation is also a, let the user understand the value of the product, and find the product irreplaceable.
Lead after recall
Many of the products in the recall process only the user recall is considered merit, in fact, there is an important closure of the recall work can be done, that is, after the recall guide.
The user has finally returned to the product's various recalls, but if he sees a product that is no different than the one he left, his next loss is on the agenda, so how to get the lost recall user back into the retained user, and the lead job after the recall is essential.
Crazy operation: How to keep users active
Keywords: User-graded growth path herd effect fun and easy to use
There are two aspects to the active work, on the one hand, to make inactive users active, on the other hand, to maintain active user's active degree.
Active work can be combined with users to do the loss, the first to do user analysis, then user classification, according to user classification features to develop active strategy. From the point of view, the prevention of loss is to keep the user, the next step is to let the retained users become active.
Establish active measures according to user registration time
Find inactive users are now in the growth of the stage, that is, to determine whether the user is newly registered users or a long time to register the old users.
If it is not long after the registration of users, this time is the best time, will be registered to guide the process of pain cut off the icing on the cake of the function introduced at this time can be taken out to show. User patience at this time is much better than registering. Most of them for the product is still in the groping period, this time the hot, through the product function guide with him to play, let him feel that the product in the process of being cared for, rather than being free-range, for the rapid development of product awareness and loyalty plays a vital role.
If you are a long time old user, you need to do more detailed analysis, such as whether they have been active before? Is it active or is it always wandering outside the product door?
If the enthusiasm of the decline is active, it is necessary to focus on the reasons for the decline in the degree of activity. If it is due to the normal enthusiasm of the decline, you need to be based on their interests and hobbies, quickly find other fun, easy to use the function, if it is because the enthusiasm of a feature because of upgrades or updates caused by passion, you need to quickly guide to help them to familiarize with the updated product features, At the same time, it reflects whether the active degree caused by the function upgrade is a case or group phenomenon, and it should adjust the idea of product function updating according to the user's real reaction.
Simulate user Growth path
The above mentioned are through the product mechanism to guide users, educating users, to help users improve the degree of activity. Many products think this kind of way is very passive, and the user is apt to dislike, then we try another way: with the user with the user.
Allow users to increase user activity
Active users are a bunch of products have a more comprehensive understanding, and has a good loyalty of users. So how do they pass on the knowledge and loyalty of their products to inactive users?
Electricity quotient is a very good example.
We know that people have a herd mentality, when shopping to find several homogeneous stores, we always choose to have customers, or more customers shop to buy. So often there are clearly two prices are the same, but a shopkeeper busy, another shop before the empty, is because of the relationship between the herd, also can be seen as a demonstration of the role.
How many times have you been "in the herd"?
But the electricity trader moves this kind of transaction behavior to the Internet, the credit and the appraisal system arises. The user likes to choose the merchant according to the store's praise rate and the credit degree. Also because of this, we see the typical electric Shang treasure shop Brush Drill and brush praise as crazy.
In addition, the electric business industry data Mining also embodies the herd mentality. Amazon is the first to do user purchase behavior Data Mining Enterprises, this purchase behavior data Mining will help you find the goods you may like, users do not think this is a commercial push, but based on the real user data intimate screening.
Everybody likes this one.
Information products and entertainment products are generally used to achieve the role of demonstration through business promotion. For example, the foreign claims that the product reached * million users at the same time online, or let celebrities in public to show their products. If it is a game product, you can also use such as a team to play strange and other ways to guide users and users to create a relationship between the role of demonstration.
SNS Community products have a good mentoring gene in active user-driven inactive users. Because they connect users through relationships, there are always inactive users in every active user's circle, and there are always active users in every inactive user's circle of communication. Therefore, through intentional guidance, active users in the use of products in the process can spontaneously use the fun of the product, methods to pass to inactive users, thus forming a relatively natural chain of drive.
In fact, SNS products in the commercial promotion, but also the use of the user's herd mentality, coupled with the role of celebrity demonstrations. For example, Sohu Weibo: to Sohu Weibo look at me, I am * * *, and so on, such as Tencent Micro-letter and UC Browser can not wait to report their own billion-level users-hundreds of millions of people are using, how can I out of it!
I'm Li Na, I'm weibo.com.
Maintain active user Activity
Keep active users active, the basic principle is to minimize the operational level of interruption. These active users are not only familiar with the product, but also enjoy, they do not want to be destroyed such interest.
Therefore, we need to consider whether there is any interruption to this active user when we develop some product operations strategies. Now many products launch new features, keen to give users strong guidance, such as bubbles, such as novice guide. Log in today, to remind the new features, if you do not open to see the details, perhaps the next login will remind, this behavior I can only say to dig my own grave. During the function promotion, active users are repeatedly harassed, to the end can only close the page/client, the whole world clean ...
Are you going to check it out?
Therefore, this is actually a need to think back to the problem, active users as far as possible not to disturb, inactive users as far as possible to care, can not simply follow the product function to promote the cycle of the user to do a one-size-fits-all operation strategy.
Core User Operations
Core users include: Users who are able to make constructive comments on products in the course of their use, who have the obvious popularity or the right to speak of the old users, the user has a deep understanding of the product, can provide obvious help for product improvement. Such users, we should keep in close contact with them, and provide the interests of the minority exclusive.
At the beginning of the product launch, such a core user is very rare, but this time is precisely the most need for the core users, the more such users more precious, their user experience the most authentic, their promotion most powerful. Many people think that Sina's blog and micro-Bo take the star effect, but these users have a professional and a high degree of vision, not only to quickly feedback product experience, but also quickly find a foothold in the early product has an indelible role.
Summary
In fact, to do user operations, is basically to make users feel that this product can not only use, but also good. From the strategic point of view, opened the source, the section has a flow, sanity active, only to be considered excellent user operations.
From the product point of view, to attract users in the initial registration, fun, fresh, is an essential element. However, in order to develop in the subsequent development of thick thin hair, but also need this product has practical, so-called irreplaceable, refers to the availability of irreplaceable. Attract users, you want fun, keep users, you need to use.
Attach a strong user operation information map:
Author: Jin Pu (@ Little Yan), Sina Weibo operations manager, focused on user operations.