Avoid Delegation Warnings When Using If() Inside Filter() in Power Apps

Instead of placing the conditional logic inside the filter predicate, move the decision-making outside and apply separate Filter() statements.

If(
    !IsBlank(DatePicker.SelectedDate),
    Filter(dataSource, Created <= Today() - 15),
    Filter(dataSource, Created <= DatePicker.SelectedDate)
)

Why This Works

Each branch contains a simple Filter() expression. The filter condition remains straightforward and delegable. Power Apps can translate each query more effectively for supported data sources.

Better Approach 2: Precompute the Value and Then Filter

Another clean approach is to calculate the date value first and store it in a variable.

Set(
    varCutoffDate,
    If(
        !IsBlank(DatePicker.SelectedDate),
        Today() - 15,
        DatePicker.SelectedDate
    )
);
Filter(
    dataSource,
    Created <= varCutoffDate
)

Why This Works

• The If() function executes once outside the query.
• The Filter() predicate remains a simple comparison.
• The variable can be reused elsewhere in the application.
• The formula is easier to read and maintain.

When Should You Use This Pattern?

This approach is particularly useful when:

• Working with large SharePoint lists.
• Using SQL Server or other delegable data sources.
• Reusing the calculated value across multiple controls or screens.

In many cases, the variable-based approach is the most maintainable solution because it separates business logic from data filtering logic.

Key Takeaways

• ❌ Avoid using If() (or other non-delegable functions) directly inside Filter().
• ✅ Keep the Filter() predicate as a simple, delegable comparison.
• ✅ Move conditional logic outside Filter() whenever possible.
• ✅ Use an outer If() or a precomputed variable to maintain delegation.
• ✅ Verify delegation behavior when working with large datasets.

Summary

A common cause of delegation warnings in Power Apps is placing conditional logic such as If() directly inside a Filter() function. Although the individual comparisons may be delegable, the combined expression often is not. By moving the decision-making outside the filter or precomputing values in a variable, you can keep your queries delegable, improve performance, and ensure accurate results when working with large data sources.