categorize duration for selected cases

Overview

The Categorize Duration for Selected Cases enrichment transforms continuous duration measurements into actionable performance categories for specific subsets of your cases. This powerful tool enables you to classify case durations or any TimeSpan attribute into five distinct performance bands: Fast, Normal, Slow, Extreme, and Negative. What sets this enrichment apart is its ability to apply categorization selectively to filtered case populations, making it ideal for comparative performance analysis across different vendors, customers, regions, or any other business dimension.

This enrichment uses intelligent statistical analysis to automatically calculate category boundaries based on percentile distributions (20th, 80th, and 90th percentiles) of the filtered data, or you can define custom thresholds that align with your business service level agreements. The result is a new text attribute that enables instant performance segmentation in visualizations, filters, and downstream analysis workflows.

By combining duration categorization with case filtering, you can answer critical business questions like "How do invoice processing times for our top vendors compare to smaller suppliers?" or "Are high-value orders processed faster than standard orders?" This enrichment is essential for performance benchmarking, SLA compliance tracking, and identifying process bottlenecks across different business segments.

Common Uses

Categorize purchase order approval durations for specific vendors to identify which suppliers experience delays
Classify invoice payment cycle times for high-value customers versus standard customers to ensure priority processing
Segment claim resolution times for specific insurance products to identify underperforming claim types
Analyze loan application processing durations for different risk categories to optimize approval workflows
Compare order fulfillment lead times across different distribution centers or geographic regions
Evaluate patient treatment durations for specific medical conditions to establish performance benchmarks
Track manufacturing cycle times for different product families to identify production bottlenecks

Settings

Filters: Define which cases should be included in the categorization. Use filters to select specific subsets of your event log, such as particular vendors, customers, products, regions, or any other case attribute. The categorization and statistical calculations will apply only to cases matching these filter criteria. When filters are specified, the default duration thresholds (Fast, Normal, Slow) are automatically calculated based only on the filtered case population. This allows you to create performance categories tailored to specific business segments. If no filters are applied, all cases in the dataset will be categorized.

Attribute Name: Select the duration attribute you want to categorize. This must be a TimeSpan type attribute, such as "Case Duration," "Duration Between Two Activities," or any other enrichment that produces duration values. The dropdown menu displays only valid TimeSpan attributes from your event log. Common choices include overall case duration, time between specific milestones, or activity-level duration measurements.

New Attribute Name: Specify the name for the new performance category attribute that will be created. By default, this is set to "[Attribute Name] - Category" (e.g., "Case Duration - Category"). This new text attribute will contain one of five values: "Fast," "Normal," "Slow," "Extreme," or "Negative." Choose a descriptive name that clearly indicates what is being categorized and for which population. The new attribute can be used immediately in filters, charts, and other enrichments.

Fast Duration: Set the upper threshold for the "Fast" performance category. Cases with durations less than or equal to this value will be classified as "Fast." When you select an attribute and optionally apply filters, this value is automatically set to the 20th percentile of the filtered duration distribution, meaning 20% of cases fall into the Fast category. You can override this with a custom value that aligns with your business SLA or performance targets. The value is specified in the selected Duration Unit.

Normal Duration: Set the upper threshold for the "Normal" performance category. Cases with durations greater than Fast but less than or equal to this value will be classified as "Normal." The default is automatically set to the 80th percentile of the filtered duration distribution, meaning 60% of cases (between the 20th and 80th percentiles) fall into the Normal category. Adjust this threshold to match your organization's definition of acceptable performance. The value is specified in the selected Duration Unit.

Slow Duration: Set the upper threshold for the "Slow" performance category. Cases with durations greater than Normal but less than or equal to this value will be classified as "Slow." The default is automatically set to the 90th percentile of the filtered duration distribution, meaning 10% of cases (between the 80th and 90th percentiles) fall into the Slow category. Cases exceeding this threshold are classified as "Extreme." This setting helps you identify performance outliers that may require attention. The value is specified in the selected Duration Unit.

Duration Unit: Select the time unit for specifying the Fast, Normal, and Slow duration thresholds. Available options are Days, Hours, Minutes, and Seconds. The default is Days. When you change the unit, the displayed threshold values automatically convert to the new unit while preserving the underlying TimeSpan values. Choose the unit that makes the most sense for your process timeframes - use Days for long-running processes like procurement or loan origination, Hours for order fulfillment, and Minutes or Seconds for operational processes.

Reset Button: Click this button to recalculate the default duration thresholds based on the current filtered case population. This is useful when you've manually adjusted the thresholds but want to return to the statistically-derived defaults (20th, 80th, and 90th percentiles). The reset operation uses only the cases matching your current filter criteria, ensuring the defaults are relevant to the selected population.

Examples

Example 1: Purchase Order Approval Performance by Vendor Tier

Scenario: A procurement team wants to analyze purchase order approval times specifically for their top-tier strategic vendors (those with annual spend over $1 million). They need to categorize approval durations to identify which strategic vendors experience exceptional, normal, or problematic processing times, enabling them to prioritize vendor relationship improvements.

Settings:

Filters: Vendor Tier = "Strategic" (cases where annual vendor spend > $1,000,000)
Attribute Name: PO Approval Duration
New Attribute Name: Strategic Vendor PO Performance
Fast Duration: 2.5 (Days)
Normal Duration: 5.0 (Days)
Slow Duration: 7.0 (Days)
Duration Unit: Days

Output:

The enrichment creates a new case attribute called "Strategic Vendor PO Performance" with the following values:

Case ID	Vendor	PO Approval Duration	Strategic Vendor PO Performance
PO-1001	Acme Industrial	1.8 days	Fast
PO-1002	Global Supplies	4.2 days	Normal
PO-1003	Premier Materials	6.5 days	Slow
PO-1004	Mega Corp	9.2 days	Extreme
PO-1005	FastTrack Vendors	3.8 days	Normal

The categorization applies only to strategic vendors. Cases for non-strategic vendors are not categorized (the new attribute remains empty for those cases). The thresholds were automatically calculated based on the duration distribution of strategic vendor POs: 20% of strategic vendor POs are approved within 2.5 days (Fast), 80% within 5 days (Normal threshold), and 90% within 7 days (Slow threshold).

Insights: The procurement team discovers that 12% of strategic vendor POs fall into the Extreme category (over 7 days), primarily concentrated with three specific vendors. This triggers a root cause analysis revealing that these vendors use non-standard requisition forms. By addressing this issue and standardizing the intake process, the team reduces extreme-duration POs by 75% within three months.

Example 2: High-Value Invoice Payment Prioritization

Scenario: An accounts payable department must ensure that invoices over $50,000 are paid faster than standard invoices to maintain critical vendor relationships. They want to categorize payment cycle times specifically for high-value invoices to monitor compliance with their internal SLA of 10-day payment for large invoices.

Settings:

Filters: Invoice Amount > 50000
Attribute Name: Invoice to Payment Duration
New Attribute Name: High-Value Payment Performance
Fast Duration: 5 (Days)
Normal Duration: 10 (Days)
Slow Duration: 15 (Days)
Duration Unit: Days

Output:

The enrichment creates a new case attribute called "High-Value Payment Performance" for high-value invoices:

Case ID	Invoice Amount	Invoice to Payment Duration	High-Value Payment Performance
INV-5001	$125,000	4.2 days	Fast
INV-5002	$87,500	9.8 days	Normal
INV-5003	$95,000	13.5 days	Slow
INV-5004	$150,000	18.7 days	Extreme
INV-5005	$72,000	6.1 days	Normal

The Fast category (under 5 days) represents exceptional processing. Normal (5-10 days) meets the SLA. Slow (10-15 days) indicates SLA violations that need attention. Extreme (over 15 days) represents critical delays requiring immediate investigation.

Insights: By filtering the process map to show only Extreme performance cases, the AP team identifies that high-value invoices lacking purchase order references account for 78% of payment delays. They implement an automated matching enhancement and reduce Extreme-category payments from 15% to 3% of high-value invoices within two months.

Example 3: Regional Claim Processing Performance

Scenario: An insurance company wants to categorize claim processing times for different regions to identify areas needing process improvement or additional resources.

Settings:

Filter Cases: Region = "Northeast" AND Claim_Type = "Auto"
Attribute Name: Claim_Submission_To_Payout
New Attribute Name: Northeast_Auto_Claim_Category
Fast Duration: 72 hours (auto-calculated)
Normal Duration: 168 hours (auto-calculated)
Slow Duration: 240 hours (auto-calculated)
Duration Unit: Hours

Output: Creates "Northeast_Auto_Claim_Category" with distribution:

"Fast" - Claims resolved within 3 days
"Normal" - Claims resolved within 3-7 days
"Slow" - Claims resolved within 7-10 days
"Extreme" - Claims taking more than 10 days

Insights: Comparing Northeast performance to other regions shows that the Northeast has significantly more claims in the Slow and Extreme categories, leading to the deployment of additional claims adjusters to that region.

Example 4: Manufacturing Batch Quality Categories

Scenario: A pharmaceutical manufacturer wants to categorize production cycle times for specific drug batches that require additional quality controls.

Settings:

Filter Cases: Product_Category = "Controlled Substance" AND Batch_Size > 1000
Attribute Name: Manufacturing_Cycle_Time
New Attribute Name: Controlled_Batch_Performance
Fast Duration: 4.5 hours (auto-calculated)
Normal Duration: 8 hours (auto-calculated)
Slow Duration: 10 hours (auto-calculated)
Duration Unit: Hours

Output: New attribute showing:

"Fast" - Batches completed in under 4.5 hours while maintaining quality
"Normal" - Standard production time of 4.5-8 hours
"Slow" - Extended production time of 8-10 hours
"Extreme" - Batches requiring more than 10 hours, indicating potential issues

Insights: Analysis shows that batches in the Extreme category have a 3x higher rate of quality issues, leading to new protocols for batches exceeding 10 hours of production time.

Example 5: IT Ticket Resolution by Priority

Scenario: An IT service desk wants to create performance categories for high-priority tickets from VIP users to ensure SLA compliance.

Settings:

Filter Cases: Ticket_Priority = "P1" AND User_Category = "VIP"
Attribute Name: Ticket_Resolution_Duration
New Attribute Name: VIP_P1_Resolution_Category
Fast Duration: 30 minutes (manually set per SLA)
Normal Duration: 2 hours (manually set per SLA)
Slow Duration: 4 hours (manually set per SLA)
Duration Unit: Minutes

Output: Categories created:

"Fast" - P1 VIP tickets resolved within 30 minutes (target: 50%)
"Normal" - Resolution within 30 minutes to 2 hours (target: 40%)
"Slow" - Resolution within 2-4 hours (acceptable: 8%)
"Extreme" - Resolution exceeding 4 hours (must be < 2%)

Insights: Current performance shows only 35% of P1 VIP tickets in the Fast category, below the 50% target. The service desk implements automated escalation for VIP tickets approaching the 30-minute threshold.

Output

When this enrichment is executed, it creates a new case attribute with the following characteristics:

Attribute Details:

Data Type: String (text)
Attribute Type: Performance (automatically marked for performance analysis features)
Scope: Case-level attribute (one value per case)

Category Values: The new attribute will contain one of five possible values:

"Fast": Duration is less than or equal to the Fast threshold
"Normal": Duration is between Fast and Normal thresholds
"Slow": Duration is between Normal and Slow thresholds
"Extreme": Duration exceeds the Slow threshold
"Negative": Duration is less than zero (data quality issue)

Null Handling: Cases without a value in the source duration attribute will have null in the category attribute. This includes:

Cases filtered out by your selection criteria
Cases missing the duration value
Cases with invalid duration data

Using the Output: The categorized attribute can be used in:

Performance matrix visualizations to show distribution across categories
Case filters to focus on slow or extreme cases requiring attention
Conformance checking to identify cases violating SLA categories
Comparative analysis between different filtered groups
Dashboard KPIs showing percentage of cases in each category
Root cause analysis to understand what drives extreme performance

Integration with Other Features:

Use in calculated attributes to create scoring systems
Combine with other enrichments for multi-dimensional performance analysis
Export to external systems for SLA reporting
Use as input for predictive models to forecast performance categories

Overview

Common Uses

Settings

Examples

Example 1: Purchase Order Approval Performance by Vendor Tier

Example 2: High-Value Invoice Payment Prioritization

Example 3: Regional Claim Processing Performance

Example 4: Manufacturing Batch Quality Categories

Example 5: IT Ticket Resolution by Priority

Output

See Also