It turns out that there are so many ways to do a small push?

Do you still remember the 201-pound operation brother mentioned in my last article "Product Manager Demand Paradigm"? He gave me a profound philosophical question today "Age, are we brothers?". After looking at his fist as big as a casserole, I decided to go with the flow: "It goes without saying that of course we are brothers."

Operation brother: "Well, you can help me write a push strategy system for the car owner APP. If you write it well, we will be brothers."

The functions and principles of push****

Active push strategy

Automatic push strategy

push**** statistics

I believe that all of us receive a lot of push messages every day, so much so that we have to turn off the message notification function of most apps.

So the question is, why are APP operators so keen on pushing messages to us (users)?

The answer is simple: of course, wake us up and use their APP.

If Push is used well, it can certainly effectively wake up users. However, if it is not done well, it will not only seriously disturb users, but also cause negative consequences: at least it can turn off message notifications, and at worst it can uninstall the APP directly.

Therefore, push must follow three major principles

Speak to “users”

Triggered by “users”

and “users” About

Speak to the user: Use the information the user has previously left on the product to push relevant messages to him.

Triggered by "user": By mining the user's behavior sequence, push content linked to a certain behavior point

Related to "user": The content pushed this time is really Contact with the user

For example:

The information left by Agger in the car owner APP is about daily use of the car in Suzhou. Then the operation can push the relevant content of the Suzhou Che Club to me and invite Ag to join the Suzhou Che Club. This kind of information is relevant to me, and I may be interested in clicking on it.

If the information pushed is about the National Car Club and I am asked to find it by myself, I will probably be too lazy to find it.

Push is actually divided into two situations: active push and triggered push. Active push means that operations push messages to users based on their actual situation.

The main process is as shown in the figure: Business goal-select the content to be sent-select the group of people to send-select the sending channel-send time setting-push data statistics

Business goal: every push, It should have its own purpose. Only when business goals are clear can operational means be divided. Of course the goal won't be present in the product.

Select the content to send: There are many types of content to send, the common ones are "coupons, H5 links, pictures and texts, texts, pictures and texts, etc."

Selection of the people to send: different contents , should be pushed to different users, so that every push will not disturb all users. Specific methods include: tag selection, crowd clustering, basic information, user behavior statistics, etc.

Sending channel selection: According to the product type, there are generally APPs, text messages, mini programs, public accounts, etc.

Send time setting: This is for convenience of operation and is divided into two types: instant and scheduled. The specific time is recommended to be set based on the analysis of target user active time data.

For example:

This time we are implementing a push strategy for the APP for car owners, so based on the group clustering: (1) To novice car owners, push the familiar APP gameplay and car usage. The information is appropriate; (2) For car owners in the growth stage, it is correct to push behavioral guidance content to guide car owners in the mature stage to find their own value points in the app community; (3)...

Automatic push again It is called scene triggered push.

Mainly because the user triggers the trigger of this scene, we automatically push messages to motivate users' behavior in a timely manner and better satisfy users.

The main process is: scene confirmation-trigger confirmation-push content confirmation-push message time confirmation

Scenario confirmation: different products will have different scenarios, this requires a little bit of us Add and improve scenes.

Trigger confirmation: In the scenario, we need to set the trigger reasonably. Only when the user triggers the trigger, we will enter the next stage;

Push content confirmation: push Content, including user story maps, coupons, maintenance discount coupons, mall discount coupons, etc., cannot be designed for different products.

Push message time confirmation: Because it is an automatic push, in theory, the content can be pushed directly at any time to users, but actually we need to design the push time according to the actual scenario

Give an example:

Xiaopeng Motors users will receive the payment at 12 noon on the first anniversary of their car purchase. Go to one: User Story Map. Car owners will feel very considerate.

Push**** data statistics

By 2020, basically everyone has a common understanding: data-driven. That is to say, all functions should have statistical significance to assist the operation review and push behavior.

Generally divided into the following categories:

(1) Statistics of successful sending data

(2) Statistics of failed sending data: These two items must be coordinated with the channels. Look, it mainly depends on the smoothness of the push channel

(3) Behavioral click data statistics: you can know the quality of the push title by whether it is clicked

(4) Business conversion rate: Pass Business conversion rate, you can know the realization of business goals

I made the above content into a mind map and gave it to the operation manager.

The operation said: "From today on, Ag, you are my brother."

Reference materials:

[1] Practical data analysis, Zhang Tao , 2019.03

As we all know, product managers are a role that everyone shouts about in many companies. But Age rarely worries about operations/UI/R&D being violent to me, because Age is 180cm tall and weighs 180cm, has dancing chest muscles, is silly, big and thick, and looks more like a nightclub security guard than a product manager.

So I often make requests to UI just like the uncle in the picture: the products I want should be rich in color and have strong contrast.

Everything came to an abrupt end when I came to this new company. The operator here is a brother with a big casserole fist. The UI here is 2.2 meters tall and weighs 220.

So I decided to hide this picture and wait until the next company to post it.