The times are so affected by tools. As a product manager, we will also be affected by tools, including ink knives, nails, blue lakes, graphite, etc. More and more Mexico Phone Number new tool changes are quietly changing of this industry.

Of course, the protagonist of this article: MTA (Tencent Mobile Analytics), will continue to undergo new changes, exert its mission value as a tool, and bring new data-driven methods to the entire mobile Internet industry.

Data-driven products/operations will never be an established goal, but a sustainable process. During this process, the tools we use for data monitoring and analysis will directly affect our daily work. Influence our judgment on the direction of the product and influence our decision-making.

Ultimately, data tools will influence whether our product succeeds or fails.

Mexico Phone Number
Mexico Phone Number

In fact, I have great expectations for the topics that I will discuss next. If I can enter the iterative schedule of the MTA, I will be very fortunate to be an audience user of such a product.

What I’m looking forward to is a small change in ” active days “.

1. Overview of Active Days

Active days is an early function of MTA. The specific launch time is unknown. In my memory, this module already existed in 2017.

In simple terms, active days is a data object that records the user’s login days in the statistical period. We believe that within the period, the more active days, the stronger the user’s stickiness, and the lower the active days, the weaker the user’s stickiness.

In the Demo data provided by MTA, we found that in the 7-day cycle from 2018-5-25 to 2018-6-1, on the day of 5-29, the proportion of users who logged in for two days accounted for 11%, and some 5 days login

5% of users with behavior

(Because it is Demo data, or because of the design problem of this module, the data is difficult to interpret. The above data is only for the opinion of the article, and does not represent the official data interpretation.)

We can make a simple interpretation through the Demo data. The user stickiness of this product is not very strong. High-frequency users only account for 5% of active users, while users who log in for only two days within a week account for 11%. .

In other words, it will be easier to explain my point of view. Low-frequency users are infinitely close to losing users. Among the active users of the case object on May 29, 11% of users are about to be lost.

Under normal circumstances, those who like it will like it more, and those who don’t like it will dislike it even more. Normal data is never an instantaneous change, it is a continuous trend, and it is a process from point A to point B.

(Special events such as activities, server crashes, business exceptions, etc. can cause instantaneous changes in data, but this change is not normal, but triggered by events.)

This process is precisely one of the core values ​​that the data feeds back to us. Under normal conditions, people cannot suddenly change from liking to disliking. A user who has been using for 7 consecutive days cannot suddenly be lost. The process of changing the number of active days is precisely the best reflection of the change in user stickiness.

With the help of active days, we can judge the stickiness of users and the health of products, which is more beneficial for us to create a product that is deeply loved by users.

I have built such a product. In the 30-day observation period, more than 50% of users have active days for more than 25 days, and more than 70% of users have active days for more than 20 days.

This data shows that my users are very satisfied with this product, and they are willing to make this product a part of their daily lives, just like WeChat, integrated into our lives. And in the absence of emergencies, there will be a long security period, and users will not be lost too quickly.

If more than 50% of the active days are under 5 days, I may be anxious because I am about to lose more than half of my users.

2. Explanation of the concept of “activity” and application scenarios

I will call the data indicator that judges user stickiness by the number of active days as “user activity” . Based on the active days and supplemented by weighted points, the user activity points are obtained , and then the activity points are graded to obtain the user activity level .




Leave a Reply

Your email address will not be published. Required fields are marked *