Creating Windows Filters – Best Practice

1. Purpose

Filters in application windows allow users to efficiently search, restrict, and analyze data.
Well-designed filters improve:

• Usability
• Performance
• Data quality
• Maintainability

This document describes best practices for defining and implementing filters in ERP or metadata-driven application windows.

2. General Principles

2.1 Business-Oriented Design

Filters must reflect how users think about their data.

✅ Good Examples

• Date (Document Date)
• Business Partner
• No. (Document No.)
• Document Type

❌ Avoid

• Technical IDs (e.g., C_Order_ID)
• Internal flags not relevant to users
• Backend-only fields

2.2 Prefer Indexed Columns

Filters should primarily use:

• Indexed columns
• Frequently searched columns
• High-selectivity fields

Avoid filtering on:

• Large text fields
• Non-indexed calculated columns
• Function-based or subquery-heavy expressions

Reason: Filters must not degrade system performance.

2.3 Use the Correct Reference Type

Do not use generic string fields if a structured reference exists.

3. Functional Best Practices

3.1 Default Filters

Default filters should:

• Reflect the most common business scenario
• Not hide important data unintentionally
• Be transparent to users

Example Defaults

• IsActive = Y
• DocStatus != Closed

Avoid overly restrictive defaults that confuse users.

3.2 Date Filters

Best practices for date filtering:

• Default date should be document’s main date.
• Provide “From” / “To” range capability

Good

• Date Ordered From / To
• Delivery Date From / To

Avoid

• Fixed equality date filters
• Hardcoded date conditions in window logic

3.3 Boolean Filters (Yes/No)

Use boolean filters only when they:

• Represent meaningful business distinctions
• Significantly reduce result sets

Examples

• Is Vendor
• Is Sales Order
• Processed
• Posted

Avoid cluttering the filter panel with too many technical flags.

4. Technical Best Practices

4.1 Keep Filter Logic Simple

Avoid:

• Complex SQL subqueries
• Heavy database functions
• Cross-table dependencies when possible

Simple logic improves performance and maintainability.

4.2 Performance Considerations

For large tables:

• Ensure filtered columns are indexed
• Test with production-like data volumes
• Avoid filtering on calculated or virtual columns
• Avoid full table scans

Performance testing is mandatory for transactional windows.

5. User Experience Guidelines

5.1 Order Filters by Importance

Place most frequently used filters at the top:

1. Document No
2. Business Partner
3. Date
4. Status
5. Organization

5.2 Avoid Overloading the Filter Panel

Too many filters:

• Reduce clarity
• Increase maintenance effort
• Confuse users

Guideline:

• Include essential filters only
• Hide technical or rarely used fields

6. Testing Checklist

Before releasing a window filter:

• Does it reflect a real business use case?
• Is the column indexed?
• Does it perform well with large datasets?
• Does it respect role permissions?
• Does it behave correctly across organizations?
• Are default values safe and understandable?
• Is naming clear and business-friendly?

7. Summary

Good filters are:

• Business-oriented
• Performance-aware
• Security-compliant
• Simple and maintainable
• Focused on user needs

Well-designed filters significantly improve system usability and data transparency.