To power diverse ministry insights, Trends AI uses several data models. Whether you’re working with AI Analysis or the Chart builder, picking the right model for the task is the first step to gaining great insights.
For instructions on how to choose a data source for your Chart, please review this article.
Foundational Concepts
Data grain
Each model has a unique set of underlying data allowing you to explore your ministry data in complex ways. They also rely on a unique data grain, which is the specific shape and level of granularity of data.
Example: The Giving model is structured around individual transactions, while the People model is built around individual profiles.
Understanding the data grain is important because it determines how you can filter, group, and analyze your data in meaningful ways.
One model per chart
While a dashboard (board) can include multiple charts side by side, each individual chart can only use one data model at a time.
If you want to compare different datasets (e.g., giving trends vs. attendance trends), simply create separate charts for each and place them together on your board.
Imported tables
You can upload your own data to create a custom table, which functions as a unique data model.
An imported table functions as a unique table, so you aren't directly combining datasets within a single chart. However, this makes it easy to show Subsplash data and external data on the same board to circumvent the one-model-per-chart limit
Exploring your data models
AI Analysis
AI Analysis helps you quickly understand what questions you can ask based on your data. Start with a role-based prompt:
“I’m a bookkeeper at a church. What questions can I ask using the Giving model?”
You’ll receive suggested questions tailored to your role and available data.
View column information
Within the chart builder, you can select any column to view its description in the Column Information panel. This helps clarify:
What the field represents
How it should be used in analysis
Whether it applies to filtering, grouping, or calculations
Giving Model
Provides a detailed view of all donation activity. It helps you understand who is giving, gross/net amounts, fund distribution, and how giving methods change over time.
Primary Data Grain: Subsplash transactions (requires Subsplash Giving or imported gifts).
Key Fields:
Amount: Gross gift amount before fees.
Fund Name: Designated fund or sub-fund (e.g., "Missions").
Transaction Date: When the gift was given.
End User UUID: Unique donor identifier.
Instrument Type: Payment method (Card, ACH, etc.).
Is Recurring: Indicates if the gift came from a recurring schedule.
Use Is Successful to exclude failed transactions. Use Is Tax Deductible to filter out non-donation activity. Use Is Non Traditional to identify non-standard giving types.
Example Questions:
How is giving trending this year compared to last year?
Who are our top 10 givers this month?
What percentage of giving goes to each fund?
Recurring Gifts Model
Focuses on scheduled giving activity rather than individual transactions. Useful for forecasting and donor retention analysis.
Primary Data Grain: Recurring Gift schedules.
Key Fields:
Amount: Current recurring gift amount.
Is Active: Whether the schedule is currently active.
Interval: Frequency (Weekly, Monthly, etc.).
Next Gift Timestamp: Next scheduled gift date.
Created At: When the schedule was created.
This model reflects schedules only, not historical transactions. For past gift activity, use the Giving model.
Example Questions:
What is the trend in recurring gift creation over time?
What is the expected total recurring giving next month?
How many donors currently have active recurring gifts?
People Model
Includes all profiles in your system to help you understand community structure, demographics, and growth.
Primary Data Grain: Subsplash Profiles.
Key Fields:
Profile UUID: Unique identifier for each profile.
Profile Created At: Vital for tracking growth trends.
Profile Created Source: How the person entered the system (e.g., app sign-up).
Membership Status Name: (e.g., Member, Guest).
Total Profile Gifts: Count of transactions made by this profile.
Group Name: Groups the person belongs to.
Be aware that imported data or duplicate profiles may inflate numbers. Data like Group Name and Total Profile Gifts overlap with other models; for deeper analysis of those specific areas, use the Giving or Group models.
Example Questions:
How many new profiles are created each month?
What is our age and gender distribution?
What percentage of members have given financially?
Event Attendance Model
Combines check-in data and headcount data to analyze engagement across events and services.Th
Primary Data Grain: Subsplash Events (requires Subsplash Events and Check-in; Legacy check-in not supported).
Key Fields:
Event Title: Name of the event.
Event Start At UTC: Event start date/time.
Total Event Attendance: Sum of headcount and check-ins.
Check In Count: Individuals checked in via the app.
Session Headcount: Manually entered anonymous headcount.
Total Attendance = Session Headcount + Check In Count. Specify the source if you only need one. Filter by Event Title to isolate specific services or events.
Example Questions:
What is our total attendance over time?
Which individuals are checking in regularly over the past 12 months?
What days of the week have the highest attendance?
Group Models
Group Membership
Focuses on who is currently connected to groups and the overall group structure.
Primary Data Grain: Groups (requires Subsplash Groups).
Key Fields:
Group Name / Group Type: Useful for analyzing popular group styles.
Group Membership Count: Total count of current members.
Gender / Date of Birth: Demographic info of group members.
Filter on Group Name to exclude irrelevant or inactive groups. Group Membership Count is a current snapshot and cannot show growth over time.
Example Questions:
Which group types are the most popular?
How many individuals are members of multiple groups?
How do public groups compare to private groups in membership?
Group Attendance
Tracks participation in group events as recorded by Group Managers.
Primary Data Grain: Group Events.
Key Fields:
Group Name: Specific group associated with the event.
Start At UTC: Date/time of the event.
Attendance Count: Total of individual and headcount numbers.
Headcount: Exclusively anonymous headcount.
Specify Group Name when you want to exclude specific groups from your queries.
Example Questions:
How is group attendance trending over time?
Which groups might need support due to declining attendance?
Which people are regularly attending multiple small groups?