metric¶

Overview¶

Metrics is a function that allows you to save and reuse the settings of frequently used indicators in the Create > Chart menu.

By defining indicators that can be commonly used throughout the company as metrics, indicators from the same source can be exposed with the same numbers. If a metric's criteria changes, you can edit just one metric and it will automatically be reflected in all charts and dashboards that use that metric.

What can you do?¶

Business/Marketing Manager¶

If you define a team common KPI as a metric, such as DAU, MAU, ARPU, etc., all members can share data on the same basis.
By classifying indicators by purpose into metric categories, you can systematically manage the analysis assets of the entire team.

Data Analyst¶

Once complex derived indicators such as ARPPU, PU and PU RATE(%) are saved as metrics, they can be immediately reused without the hassle of setting them repeatedly for each chart.
Changes to a metric definition are automatically reflected in all charts that use that metric, ensuring consistency.

Get started quickly¶

If this is your first time creating a metric, follow the steps below to create your first metric.

Click the Data > Metrics menu on the left sidebar.
Click the Register button in the top right.
Enter a metric title and category.
In the Metrics area, select the events, values, and operations you want to analyze.
Click the Save button.

Note

For detailed functions such as metric registration, exposure/non-exposure settings, and metric categories, please refer to All functions.

All Features¶

Key concepts¶

concept	Description
Metrics	Define metrics that calculate based on events (e.g. COUNT DISTINCT userId for events AU = app_login)
Built-in metrics	Metrics provided by default in the analytics system
Custom Metrics	Custom metrics defined by analytics users
Metric Category	Groups to categorize metrics by type (e.g. users, sales, sessions, gameplay analytics, etc.)

Built-in metrics¶

This is a metric provided by default in the analytics system and can be copied and modified to use as a new metric.

In order to use metrics, each event and attribute must be transmitted. For events, refer to Event log transmission, and for attributes, refer to Event attribute transmission.

Category	metric	Description	Formula (event name, aggregate, attribute, filter)
User	AU	Number of users who logged in to the game, with duplicates removed	app_login · COUNT DISTINCT · userId
	AU(deviceId)	Number of deviceIds that logged in to the game, with duplicates removed	app_login · COUNT DISTINCT · deviceId
	NU	Number of users who logged in to the game for the first time, with duplicates removed	app_login · COUNT DISTINCT · userId · newUser = 'Y'
	NU(deviceId)	Number of deviceIds that logged in to the game for the first time, with duplicates removed	app_login · COUNT DISTINCT · deviceId · newUser = 'Y'
	Login Hit	Total number of game logins	app_login · COUNT · userId
	Number of concurrent users	Number of users, with duplicates removed, included in the events that the Hive SDK automatically collects every 2 minutes while the client is running	app_session_maintain · COUNT DISTINCT · userId
	Number of Terms and Conditions	Number of deviceIds, with duplicates removed, for cases where the app is launched for the first time after download	terms_agree · COUNT DISTINCT · deviceId
Sales	In-app sales	Sum of payments made in in-app stores	purchase · SUM · price
	Purchase Quantity	Sum of users' in-app store purchases	purchase · SUM · quantity
	Number of sales	Number of receipts, with duplicates removed, issued when in-app store purchase events occur	purchase · COUNT DISTINCT · transactionId
	PU	Number of users who paid in the in-app store, with duplicates removed	purchase · COUNT DISTINCT · userId
	ARPU	In-app sales generated per user	( purchase · SUM · price ) / ( app_login · COUNT DISTINCT · userId )
	ARPU(deviceId)	In-app sales generated per deviceId	( purchase · SUM · price ) / ( app_login · COUNT DISTINCT · deviceId )
	ARPPU	Average in-app sales generated per paying user	( purchase · SUM · price ) / ( purchase · COUNT DISTINCT · userId )
	PU RATE(%)	Ratio of in-app store payment users to the number of game logged in users (%)	( (purchase · COUNT DISTINCT · userId) * 100 ) / ( app_login · COUNT DISTINCT · userId )

View metric list¶

Check all metrics registered in Analytics as a list.

Display items	Description
Metric Type	Distinguish between built-in metrics and user-defined metrics
Category	Category the metric belongs to
Metrics	Metric name (click to go to detail page)
Created date	Date and time the metric was created
Generator	User who created the metric
Last modified date	When the metric was last modified
Use or not	Toggle visibility of metrics in charts

Create Metric¶

Basic information¶

Item	Required	Description
Metric Title	Required	Name displayed on analysis screen (maximum 30 characters)
Metric Description	Select	Description of metric calculation method and purpose (maximum 300 characters)
Category	Required	Select a metric classification category created within an existing company, or create a new one by entering it directly
Result display options	Select	Decimal places / percent notation / change display settings
Metrics	Required	Formula composition with event selection + value selection + operation selection. Select a rollup or date if needed. You can create complex formulas by combining multiple items with the Add button

Metric settings¶

Metrics are defined as a combination of events, attribute values, and attribute aggregation methods.

Aggregation method	Description	Main uses
COUNT	Total number of event occurrences	Event Occurrence Frequency
COUNT DISTINCT	Number of unique users/values excluding duplicates	Number of connected users, number of paying users
SUM	Sum of attribute values	Total Sales, Total Play Time
AVG	Average of attribute values	Average payment amount, average play time
MIN	Minimum value of attribute value	Minimum payment amount
MAX	Maximum value of attribute value	Highest achieved level
MAX BY	Most recent (last) attribute value	Last level

Metric Definition Example

Metric Title	Formula (Event · Aggregation · Attribute · Filter)
AU	app_login · COUNT DISTINCT · userId
In-app sales	purchase · SUM · price
Number of sales	purchase · COUNT DISTINCT · transactionId
PU RATE(%)	( (purchase · COUNT DISTINCT · userId) * 100 ) / ( app_login · COUNT DISTINCT · userId )

A complex formula that combines the aggregated results of two events, such as PU and RATE(%), can also be defined as a single metric.

Tip

Metrics only define “what to count,” and charts determine “what to look at.” For example, if you want to check the DAU/WAU/MAU indicator, you only need to define AU. After defining one AU metric, if you change the period unit on the chart to daily/weekly/monthly, it will automatically become DAU·WAU·MAU.

Rollup (Secondary Aggregation)¶

A rollup first calculates the aggregation results at a fixed time interval and then aggregates them again by the period unit of the chart (daily/weekly/monthly, and so on). Select the rollup method, the time unit, and the interval in Rollup to the right of the aggregation method. Rollup is optional.

Rollup method	Description
MAX	Largest value among the first aggregation results by interval
MIN	Smallest value among the first aggregation results by interval
AVG	Average of the first aggregation results by interval
LAST	Latest value among the first aggregation results by interval

Example

If you define a metric as app_session_maintain · Distinct count · userId with the rollup MAX · 2 min, the metric calculates the maximum of the concurrent users counted every 2 minutes. If you view this metric in a chart with the period unit set to Daily, you can check the daily peak concurrent users.

Note

In charts, you can set a rollup only when you select an event as a measure value. To apply a rollup to a metric, define the metric with a rollup on this screen in advance.

Date Selection¶

You can specify the dates that a metric aggregates, separately from the period of the chart. Click the calendar icon to the right of the attribute, set the start date and end date to Fixed Date or Relative Date respectively, and click Apply. Date selection is optional.

Type	Description
Fixed Date	Fixed to the selected dates. The value of the selected dates is always displayed regardless of the period on the chart's [Settings] tab.
Relative Date	Specifies how many days before the date viewed in the chart. For each date viewed in the chart, the value of the date that many days earlier is displayed.

Example

If you define a metric as app_login · COUNT DISTINCT · userId and view it in a chart daily from October 1 to October 7, the result depends on the date selection as follows.

Fixed Date, September 1: Only the value of September 1 is displayed, regardless of the period viewed.
Relative Date, 1 day before: The value of the day before each date viewed, that is, the values from September 30 to October 6, is displayed.

Warning

A metric with a Fixed Date does not follow the period setting of the chart. To view trends over a period, use a metric without date selection or with a relative date.

Metric Category Management¶

Create and manage categories to classify metrics by purpose. If you don't see the category you want, you can click 'Direct Entry' to enter it using the keyboard.

Category Example

User indicators (DAU, MAU, WAU)
Sales indicators (ARPU, ARPPU, in-app sales)
Session metrics (number of sessions, play time, average number of sessions)
Game play indicators (number of content plays, amount of goods acquired, amount of goods consumed)

Controlling the display of metrics¶

If the metric type is not 'Platform', you can set whether or not to expose each metric by toggling Enabled or not in the list.
Setting OFF to Create > will hide that metric from the chart's measure selection list.

Copy and delete metrics¶

Copy: Quickly create similar metrics based on existing metrics
Delete: Delete metrics that are no longer in use.