All Products
Search
Document Center

Quick Tracking:Analysis Overview

Last Updated:Feb 10, 2026

Function Overview

Analysis in Quick Tracking is divided into two main parts: Behavior Insights and User Insights.

Behavior Insights

This feature provides statistical analysis of event-based metrics. It supports drill-down analysis with grouping and filtering conditions to help you understand how users interact with your product and identify key factors that affect metrics.

This feature helps you understand user conversion or churn across behavioral steps. You can then optimize your product or implement operational strategies to increase conversion rates and achieve business goals.

You can calculate retention based on custom start and subsequent behaviors for different business scenarios and product stages to analyze user stickiness. Use the results to adjust your strategies, guide users to discover product value, retain them, and achieve real user growth.

You can view the distribution of users for a specific event metric within defined intervals. Common analyses include user distribution by the number of event occurrences, days of product use, or a specific property value. You can use distribution analysis to better understand user characteristics.

You can observe the time intervals and data distribution between specified user behaviors. Data metrics, such as maximum, minimum, and median, reflect the user experience within the product path. You can use this information to evaluate the effectiveness of your product's feature settings.

Session analysis links individual user behaviors into a cohesive whole. This helps you perform an in-depth analysis of a user's series of actions within the product.

Path analysis uses a Sankey diagram to visualize user paths within the product. It illustrates the flow of traffic between pages. You can use path analysis to validate operational strategies and refine product design.

User Insights

Audience Insights lets you select users who share common behavioral and attribute features within a specific time period. You can then perform precise analysis and outreach to these selected users.

You can divide users by their lifecycle stage. This helps you focus on, analyze, and manage users who are at risk of churning to increase the number of active users.

Document attachments

Behavior Insights and User Insights share some common features. This section explains these common features in detail.

  • Attachment 1. Selecting event metrics

Event metrics include preset and property metrics. Preset metrics support calculations for Deduplicated Count, Count, and Count per Capita. Property metrics are divided into string and numeric types. String metrics support Deduplicated Count. Numeric metrics support calculations such as Deduplicated Count, Sum, Maximum, Minimum, Average, Lower Quartile, Median, and Upper Quartile. The following table describes these metrics.

Metric Type

Metric Name

Metric Meaning

Example

Preset Metric

Deduplicated count of device IDs

The deduplicated count of device IDs for the selected event.

Number of unique OpenIDs that accessed the miniapp.

Count of device IDs

The number of triggers for the selected event associated with a device ID.

Number of times the miniapp was launched.

Count per capita of device IDs

Count of device IDs / Deduplicated count of device IDs

Average number of launches per user for the miniapp.

Deduplicated count of account IDs

The deduplicated count of logon account IDs for the selected event.

Number of accounts that paid for orders.

Count of account IDs

The total number of times the selected event was triggered by a logged-on account.

Number of times orders were paid.

Count per capita of account IDs

Count of account IDs / Deduplicated count of account IDs

Average number of paid orders per person.

Deduplicated count of entity IDs

The deduplicated count of entity IDs for the selected event.

Number of active one-ids in a cross-terminal scenario.

Count of entity IDs

The number of triggers for the selected event with a specified entity ID

Number of active sessions in a cross-terminal scenario.

Count per capita of entity IDs

Count of entity IDs / Deduplicated count of entity IDs

Average number of active sessions per person in a cross-terminal scenario.

Property Metric

Deduplicated Count

The deduplicated count of event properties.

Number of unique products exposed.

Sum

The sum of a numeric property.

Sum of order amounts for paid orders.

Maximum

The maximum value of a numeric property.

Maximum order amount for paid orders.

Minimum

The minimum value of a numeric property.

Minimum order amount for paid orders.

Average

The arithmetic mean of a numeric property.

Average order amount for paid orders.

Lower Quartile

The 75th percentile of numeric attributes

The 25th percentile value of the order amount for paid orders.

Median

The middle value of a numeric property.

The 50th percentile value of the order amount for paid orders.

Upper Quartile

The 25th percentile of values for a numeric property

Data value for the payment of 25% of the order amount

P90

The 90th percentile value of a numeric property.

The 90th percentile value of the order amount for paid orders.

  • Attachment 2. Adding filter conditions

You can filter data using system, global, event, channel, and user properties. The following table describes these properties.

Category Name

Category Description

Details

System property

This property applies to all events and is collected by the software development kit (SDK).

A preset system property, such as the SDK version.

Global property

This property, such as the region of the Ele.me service, applies to all events and is reported by the user.

A property that you enter in the global properties section of the Collection Management platform.

Event property

This property applies to all events and represents data reported by users or collected by the system, such as the payment amount for a payment event.

A property that you enter in the event properties section of the Collection Management platform.

Channel property

This property applies to all events. You can use it to view the in-app behavioral characteristics of users from different channels and their contribution to a specific metric.

A property that you enter in the channel properties section of the Collection Management platform.

User property

This property applies to all events and is sent through a specific event or a server-side API, such as user membership information.

A property that you enter in the user properties section of the Collection Management platform.

Share reflow property

This property applies to sharing behavior events. It is a key attribution field used to identify users who return to the app through a share link. Examples include `share_id` or `utm_source=share&ref=xxx`.

A property that you enter in the user properties section of the Collection Management platform.

Different filter operators are supported based on the property type. The following table describes the available operators and their definitions.

Property Type

Filter Operator

Definition

String

Equals

Equals one or more specific values.

Does not equal

Excludes one or more specific values.

Contains

Finds values that contain specific characters.

Does not contain

Excludes values that contain specific characters.

Has no value

Finds data with no property value.

Has a value

Finds data with a property value.

Numeric

Equals

Equals one or more specific values.

Does not equal

Does not equal a specific numeric value.

Greater than

Is greater than a specific numeric value.

Greater than or equal to

Is greater than or equal to a specific numeric value.

Less than

Is less than a specific numeric value.

Less than or equal to

Is less than or equal to a specific numeric value.

Between... and...

Is between two specific numeric values.

Has no value

You can find data with no order amount.

Has a value

You can find data with an order amount.

Boolean

true

The result is true.

false

The result is false.

Has a value

Finds data that has a time value.

Has no value

Finds data that has no time value.

Timestamp

Absolute time

Equals a fixed time.

Time period

Within a specific time period.

Time relative to the current time

A point in time before or after the current time.

Time relative to the event occurrence

Based on the time an event occurred, such as the time of a click event.

Time period relative to the event occurrence

A time period relative to an event occurrence. For example, the period from when a click event first occurred to when it occurred again.

Has a value

Finds data that has a time value.

Has no value

Finds data that has no time value.

The following table shows how data is categorized and displayed when filtering for 'Has a value' or 'Has no value'.

Actual Situation

Categorization for "Has a value" / "Has no value" filter

Display when showing by property value

key:'value'

Has a value

'value'

key:''

Has no value

Empty string (preset)

key:null

Has no value

Null object (preset)

key does not exist

Has no value

Unknown (preset)