The Way of User Operation: How to dig deep into the value of push

1. Foreword
If you re not growing，then you re dying. (If the company no longer grows, then the company is dying)
-Sean Ellis, "The Growth Hacker"
This sentence is not alarmist. With the disappearance of the demographic dividend, how to grow rapidly is a problem and challenge that every company must face. Acquiring users quickly means not only an increase in the level of users, but also means that they can be "faster" in the subsequent competition.
"The smell of wine is also afraid of deep alleys." User push is an effective way to help propaganda. However, almost every company is using push. Think about the mobile phone being "bombed" by various push messages every day. In this case Users will only be harassed and cannot be converted, and they will even become disgusted with the company behind the push. Either method requires cost. "Good steel must be spent on the blade". This requires operators to avoid harassing users and pursue a push method with the highest input-output ratio (ROI). which is:
1. Find the best push channel
2. Develop the best push strategy
2. The first step: channel push
Channel push is mainly to make preliminary push attempts for the planned channels and accumulate data for subsequent analysis.
Two abilities are required:
1. Coverage capacity of existing channels
2. The responsiveness of the new channel
The ability to cover multiple existing channels allows operating students to select multiple channels for push at the same time and verify the push effect, without having to consider the access of different channels; the Internet is changing with each passing day, and the rapid response and coverage of new channels is also very important. Importantly, adding new channels often means using a new way to convert users, with lower conversion costs.
Take Aurora Unified Messaging System (UMS) as an example. It currently covers seven mainstream channels: APP, SMS, email, WeChat official account, WeChat applet, Alipay life account, and Dingding Information, and is the first to support 5G information. From the official website It is found that support for channels such as enterprise WeChat and Feishu is already under planning. Regardless of the coverage of existing channels or the response to new channels, Aurora UMS is very comprehensive.

Third and second step: Analysis and screening
With the data accumulation in the first step, it is necessary to analyze the data, find the optimal push channel, and provide data support for the next step of formulating the strategy.
The push path can generally be divided into "push, click, download, activate, register, and use". Different APP paths will increase or decrease. For example, apps that support visitor mode can skip the registration process, and the "use" of shopping apps is generally Add to the shopping cart or complete a shopping...For different nodes on the push path, draw up corresponding indicators, and find problems through the conversion between indicators.

1. Channel analysis
The index data between different channels, and the data performance under the dimensions such as time and region reflect the quality of the channel.
Take clicks as an example. Through today, this week, this month, this year's data and the chain comparison, you can get a preliminary understanding of the channel’s performance; weekly clicks, monthly clicks, and annual clicks, you can see the trend of user clicks Change, and make corresponding strategies by comparing the data of the same period. For example, the number of hits during the Golden Week of last November was very high. This year, November should be prepared in advance, increase investment, and further increase the number of hits. The user’s click volume will vary due to geographic differences. Use the click volume map to find the area with poor click volume and analyze the reason. If it is not the problem of the user audience, you can try to tilt the investment in this area. Aurora UMS currently has built-in indicators such as "send volume", "delivery volume", "click volume" and other indicators. It supports the selection of different time periods, the time granularity is hour, and it can be refreshed in real time, in line with the company's internal "download volume" and "activation volume" , "Registration", "Usage", data monitoring and analysis of the entire user path. Adding the geographic information of users or their devices can draw data maps of different indicators. As shown in the figure below, it is a visual report on the number of clicks from the two dimensions of time and region.

This picture is drawn by the author
2. User analysis
The prerequisite for which information should be pushed to which users is to know who the user is, that is, to draw user portraits through data, and to classify users according to the differences in user portraits. Take a certain content community product as an example. As can be seen from the figure below, the users are mainly males born in the 1980s. Most of them live in first-tier cities. It can be preliminarily judged that these users are in the life stage of married and dad. , Due to work reasons, they are highly sensitive to information and have strong economic capabilities. Through the combination of mobile phone brands and user reports issued by major mobile phone manufacturers in recent years, user portraits can be further refined.

This picture is drawn by the author
3. Path analysis
Every link in the user path has user churn. How to reduce these churns must first know where the user has churn, and whether the churn at this link has decreased or increased compared to the same period. The following figure shows the conversion of users between different nodes, the changes in the life cycle of users, and so on. Aurora UMS provides a "send-deliver-click" conversion rate dashboard, and also supports query life cycle, especially the "callback" function of UMS, which can call back all the information status of a message to a target to operating students. When will it pass? What channels are sent to which users, and what is the reason for the failure, to help operating students better understand the conversion on the path, so as to formulate corresponding optimization strategies.

This picture is drawn by the author
Fourth, the third step: formulate a strategy
With the accumulation of the first two steps, you can answer the two questions raised at the beginning, namely:
1. Which channel is the best?
2. What kind of push strategy is optimal?
1. Channel strategy
With channel data analysis, the next step is to formulate measurement standards for channel quality, and formulate corresponding channel strategies based on the standards. Different businesses have different measurement standards. Take conversion rates and downloads as examples. Through the distribution of different channels in the quadrants, corresponding strategies can be formulated.
High-quality channels: conversion rate and downloads are both the best performers, and investment should be increased to further expand channel performance;
Potential channel: The conversion rate is very high, and the user's choice is correct. You should find ways to simplify the download process or add some download rewards, and focus on and pull new ones;
Improve the channel: the download volume is high, but the conversion is low, you can analyze whether there are obstacles in the conversion path;
Inferior channel: The reason can be analyzed, and if the analysis cost is high, it can be discarded directly.

2. User strategy
Pushing to users instead of "harassing" users requires consideration of user habits and preferences. What kind of copywriting is more likely to impress which users when they are the problems that user strategies need to solve. Combining the portrait drawn in the second step of user analysis, formulate corresponding strategies to improve push efficiency. For example, in the two channels of SMS and WeChat official account for user group A, the conversion rate of SMS is higher, and the conversion rate will reach the peak of the day from 9 am to 10 am, so you can make the corresponding according to such user preferences. User strategy. Aurora UMS can realize the refined layering of users by "setting tags" for users, and establish mapping relationships between multiple push contents and different users, and support regular sending, message priority setting, non-disturbance and other functions to further develop strategies The ability to provide a powerful tool for simultaneously ensuring user experience and information reach rate.

3. Reissue strategy
The push environment is complex and changeable, and often a piece of information is pushed out and the effect is found to be unsatisfactory. In order to ensure the final push effect, it is necessary to send out a piece of information as soon as possible. This requires two capabilities:
1. There is a strategy template library, which can be quickly called;
2. There is a reissue strategy, which can quickly execute the called template.
The Aurora UMS template support function is quite powerful. It not only supports the template settings of individual channels such as SMS, WeChat official account, WeChat applet, etc., but also supports multi-channel universal template settings. Operation students no longer need to rush to formulate strategies for emergencies Up. Aurora UMS’s reissue strategy can automatically execute the corresponding strategy based on the return situation. If the APP channel is not clicked within ten minutes, the SMS channel will be reissued.

Five, safety assurance
Security is no small matter, especially pushing operations that involve sensitive user information. It is not only necessary to manage the people who use the push, but also to protect the user's data.
1. Permission security
According to different security levels, push messages should be reviewed by management personnel with different permissions. The permissions management of Jiguang UMS can create different roles according to the company's own organizational structure, and assign corresponding permissions to different roles. High-density-level information can only be pushed after the review is approved, and it is also supported to notify relevant personnel through in-site letters and other methods to remind you to pay attention to confidential-level information.
2. Data security
User data is the wealth of a company. In addition to industries such as banking, securities, and insurance, other industries also need to pay attention. After all, the leakage of user data may not only be obtained by competitors, but also means the lack of corporate responsibility. Aurora UMS provides private cloud deployment services to store data in the intranet, greatly improving data security.
six, summary
User push is certainly a very effective method, but it cannot be "superstitious" or obsessed with certain vanity indicators. For example, a sudden change in conversion rate during a certain period of time may not be the reason for the push, but the optimization and improvement of the product just meets To meet user needs.
Quoting the "pirate model" (AARRR model) in "Growth Hacker", user push helps users understand the company's products, and it is your product that allows users to stay and follow you.