Advanced Behavior Options
For analysis requirements involving complex user behavior data, MoEngage provides advanced analysis options. These features enable elaborate computations and structuring compared to standard count-based reports, yielding detailed results.
-
Aggregation distribution: This option lets you distribute users into buckets of aggregated event attributes. For example, you can get the distribution of users based on the minutes of songs played.
You can also configure Custom Distribution bucketing. With custom distribution, you will enter the boundaries in the From and to fields (the lower and upper limits) and an interval size to divide each distribution. This option is available only for numerical properties. -
Total events per user: This option provides an aggregated count of the total events per individual user. For example, you can get the average number of songs played per user. The aggregation types available are shown below:
-
Attribute aggregation per user: This is a two-fold aggregation where:
-
The attribute is aggregated per user.
- The attribute is aggregated again with data operations.
The result obtained will provide the arithmetic value of the attribute aggregated per user.
For example, you can know how long the least active listener played. (Minimum of the Sum of minutes)
-
When to Use Advanced Behavior Options
While standard behavioral reports are excellent for understanding counts and frequencies (For example, "How many users performed 'Song Played'?"), the Advanced Behavior Options unlock deeper, more nuanced insights when simple counts aren't enough.
Use these advanced features when your analysis requires:
- Understanding Distributions: You need to segment or understand users based on the magnitude or range of their actions, not just whether they performed an action.
- Calculating Per-User Aggregates: You need to know the average, sum, minimum, or maximum value related to an event or attribute for each individual user.
- Complex, Multi-Level Aggregations: Your question involves calculating a metric for each user first, and then performing another calculation (like finding the minimum or average) on those per-user results.
Here’s how each specific option helps solve different analytical challenges:
Aggregation Distribution
- What it does: Groups users into defined buckets based on an aggregated numerical event attribute (For example, total minutes played, total purchase value). You can use predefined or custom ranges.
-
When to use it:
- When you need to segment users based on their level of engagement or value (For example, low, medium, and high spenders).
- When you want to visualize how users are spread across different levels of activity for a specific metric.
- When understanding the distribution is more important than just the overall average or sum.
Example Use Cases
Some of the Use Cases that you can solve using this option:
- E-commerce: Identify user segments based on total spending over the last 30 days (For example, buckets of $0-$50, $51-$200, $201+).
- Media/Streaming: Understand content consumption patterns by distributing users based on 'Total Watch Time' or 'Total Songs Played' within a period (For example, 0-10 mins, 11-30 mins, 31+ mins).
- Gaming: Group players based on 'Total Levels Completed' or 'Total In-App Purchases Value.'
- Content Platforms: Segment users by 'Number of Articles Read' or 'Total Time Spent Reading.'
Total Events Per User
- What it does: Calculates an aggregate (Sum, Average, Min, Max, etc.) of the number of times an event occurred for each individual user.
-
When to use it:
- When you need to understand the frequency or volume of actions on a per-user basis.
- When you want to find the average, minimum, or maximum number of times users perform a key action.
Example Use Cases
Some of the Use Cases that you can solve using this option:
- Engagement: Calculate the 'Average Number of App Opens per User' per week.
- Feature Usage: Find the 'Maximum Number of Searches performed by any single User' last month.
- Activity Level: Determine the 'Average Number of Orders Placed per User.'
- Content Interaction: Get the 'Average Number of Videos Watched per User' daily.
Attribute Aggregation Per User
- What it does: Performs a two-step aggregation. First, it aggregates a numerical attribute for each user individually (For example, Sum of 'Purchase Amount' per user). Second, it applies another aggregation across those per-user results (For example, Minimum of the per-user sums).
-
When to use it:
- When you need insights derived from comparing individual user aggregates.
- When you want to find the extremes (For example, the user who spent the least in total, the user with the highest average session length).
- When you need to understand overall patterns based on metrics calculated at the individual user level.
Example Use Cases
Some of the Use Cases that you can solve using this option:
- Identify Least Active Users: Find the 'Minimum of the Sum of 'Session Duration'' across all users to see the lowest total engagement time. (Example: "How long did the least active listener play?")
- Understand Power User Behavior: Calculate the Maximum of the Sum of 'Levels Completed' to find the highest total level progression achieved by any single user.
- Analyze Spending Patterns: Determine the Average of the Maximum 'Single Transaction Value' per user to understand the typical highest purchase amount across your user base.
- Content Consumption Insights: Find the Average of the Sum of 'Watch Time'' per user to understand the typical total viewing duration for users.
By leveraging these advanced options, you can move beyond simple event counts to uncover sophisticated patterns in user behavior, enabling more precise segmentation, targeted campaigns, and informed product decisions.