1 How should BytePlus Growth Analysis be connected? #

The process for connecting your product to BytePlus Growth Analysis is as follows:

2 The events are all English names. Can you add Chinese remarks? #

Sure. The original English name of the event is displayed by default. You can add a Chinese "event description" to the event by clicking the "Management-Data Management" function on the navigation bar. After saving, the description content will be displayed together with the event name in the analysis function, which is convenient for understanding the meaning of the event.

3 How to check the IP of visiting users in BytePlus Growth Analysis? #

You can see through User Details the function that in the corresponding user's behavior flow, the value corresponding to the "client_ip" field is the user's IP.

4 Why does the software version number only display the large version number? #

The app_version in the SDK displays the large version number by default. If you have the need to count the small version number, you need to set the app_version_minor (that is, the second version number, the four-digit version number of the app, and the string reported by yourself can support any format. The data stream is not processed) as the small version number. It should be noted that the historical data cannot be filtered after the setting, and the statistics will start after the setting is successful.

How to merge and view 5 related events? #

• Events are OR relationships -- virtual events.

In "virtual events", you can combine multiple events in BytePlus Growth Analysis into virtual events through certain logical relationships, which is convenient for users to manage and view the combination of regular composite events.

• It's an AND relationship -- a combination metric.

In the event analysis function, you can create "composite metrics" to calculate between events and present the calculated results.
For event analysis related content, please refer to (Event Analysis) help documentation.

6 How to use analytics to win the bidding process for important changes? #

Users can use BytePlus Growth Analysis's "date annotation" function to record important changes made at certain time points (such as version releases, operational activities, online incidents, etc.), and the analysis function can be seen at a glance. Click "App Management - Date Annotation" to operate the date annotation.

How to continuously monitor and warn important data? #

In Kanban, we can keep an eye on fluctuations in key data by creating monitoring tasks. According to the monitoring rules, it can automatically determine whether the indicator fluctuates within the normal range, and trigger an alarm (email) when abnormal, without manual daily tracking and judgment, to achieve automatic monitoring.

Why is the page loading so slowly? #

Please check whether you have uploaded the number of properties under the event. Uploading too many properties under the event will cause the page to load slower. Please check whether you have set value to key.

Why can't BytePlus Growth Analysis's A/B test traffic be allocated? #

BytePlus Growth Analysis's A/B testing function has 100% traffic and does not support traffic reuse. Therefore, when the experimental traffic you open has reached the upper limit of 100%, when you start a new experiment, the traffic will not be distributed. You need to end some experiments and create them again, or contact customer service to upgrade and use the advanced product DataTester for A/B testing.

How many pieces of data can be exported by user grouping at present? #

The current limit is 200,000 pieces of data can be exported.

Why is there no data for sharing analytics in the Mini Program app? #

The use of "sharing analytics" depends on the reporting of "on_share", please confirm whether the Mini Program SDK is configured auto_report true at initialization.

12 Is data export available? #

Most charts support downloading as Excel files locally. The original data source supports exporting user and device information: including user id, public attributes, etc.; behavioral events: including events, event attributes, etc. You need to contact us to open it before use.

How are new and old users defined? #

According to whether the user event is first reported within the query time period, it is determined whether the user is a new user during this time.
For example, if the query time range is January 1, 2020 to January 3, 2020, the date granularity is day-level, and the first activation time of a user is 18:27 on January 1, 2020, then the user is counted as a new user on January 1, 2020 under the current query conditions.

14 Added some questions about new and old users after the preset event property $is_first_day (whether it is accessed on the first day) #

1. How can we distinguish between new and old users in our analysis?
   After adding the preset event attribute $is_first_day, when we analyze, we need to distinguish between new and old users by selecting "whether it's the first day". The event attribute is $is_first_day = true for new users, and $is_first_day = false for old users;
   Take event analysis as an example to illustrate the following:
   If you want to analyze the situation of new users who have visited the page, you need to select "whether it's the first day" through the filter conditions, and select "first day" as the new user; similarly, select "not first day" as the old user.;
   
   At the same time, enable grouping display by event attributes, and select "Whether it's the first day" to group by new and old users.
2. There is already a "new and old users" attribute to distinguish between new and old users. Why do you need to add the $is_first_day attribute to distinguish between new and old users? What changes have been made to the computational logic?
   The calculation logic of $is_first_day is exactly the same as that of new and old users, except that the calculation of new and old users will be affected by time granularity, while $is_first_day is only calculated by day:
   • In the judgment logic of "new and old users":
     The new user judgment logic is that the target event and the user's first event occur within the same period, otherwise it is determined to be an old user. This can lead to some problems. For example, a user is considered a new user within a 30-day cycle (e.g., the first event occurs on the 1st, and the target event occurs on the 5th, which counts as a new user), but it is considered an old user within a 1-day cycle.
   • In the judgment logic of "whether it's the first day":
     The new user decision logic is that the target event and the user's first event occur on the same day, rather than within the same cycle. When an event occurs for the first time, the DataFinder records an activation time property: first_event_time, if the event time of subsequent events is equal to first_event_time, then $is_first_day = true; otherwise, false.
3. Will the "old and new users" attribute continue to be retained?
   Considering the issue of historical data, this property platform will be retained for a certain period of time.