Process Variants

Process variants are the different execution paths that cases take through your process. Understanding variants helps you identify exceptions, compare performance between different routes, and optimize for the most common (or most important) paths.

What Are Variants?

A variant is a unique sequence of activities that cases follow through your process. Even though you might have a "standard" process, real execution often varies based on conditions, exceptions, and business rules.

Simple Example

In a loan application process, you might have:

Variant 1 (70% of cases): Standard Path

Application → Credit Check → Risk Assessment → Approval → Disbursement

Variant 2 (20% of cases): Expedited Path

Application → Quick Assessment → Approval → Disbursement

Variant 3 (10% of cases): Rejection Path

Application → Credit Check → Risk Assessment → Rejection

Each variant represents a distinct way cases move through the process.

Why Multiple Variants Exist

Variants emerge from:

• Business rules: High-value loans get extra review
• Exceptions: Problems trigger additional steps
• Manual decisions: Officers choose different paths
• Rework: Cases loop back for corrections
• Special handling: Urgent cases skip steps

Why Variants Matter

Understanding your variants reveals critical insights:

1. Identify the Happy Path

The most common variant (usually 40-70% of cases) is often your intended "standard" process. But what if it's only 30%? That suggests your process has significant variation.

2. Spot Problematic Patterns

Variants with rework loops or many steps often perform poorly:

Application → Credit Check → Risk Assessment → 
Rejected → Resubmit → Risk Assessment → Approval
(Rework variant - takes 3x longer)

3. Compare Performance

Different variants have different durations:

• Standard path: 5 days average
• Expedited path: 2 days average
• Rejection path: 3 days average (faster because abbreviated)

4. Optimize What Matters

Focus improvements on:

• High-frequency variants (affect most cases)
• Slow variants (biggest time savings potential)
• Business-critical variants (high-value cases)

Conformance Types

Flow Myna automatically categorizes variants to help you understand their nature:

Happy Path

Definition: The intended, standard execution path Characteristics:

• Usually the most common variant (highest frequency)
• Follows expected sequence
• No rework or loops
• Reasonable duration

Example:

Application → Credit Check → Risk Assessment → Approval → Disbursement
Cases: 175 (70%)
Avg Duration: 4.2 days

Anomaly

Definition: Unusual patterns that deviate significantly from normal Characteristics:

• Rare occurrence (< 5% of cases)
• Unexpected sequence of activities
• May skip expected steps
• Could indicate data issues or special handling

Example:

Application → Approval → Credit Check → Disbursement
Cases: 8 (3%)
Avg Duration: 1.5 days
Why unusual: Approval before credit check

Rework

Definition: Contains loops where cases return to previous steps Characteristics:

• Back-pointing transitions in sequence
• Same activity occurs multiple times
• Usually indicates corrections or rejections
• Typically longer duration

Example:

Application → Review → Rejected → Resubmit → Review → Approved
Cases: 25 (10%)
Avg Duration: 9.8 days
Loop: Review appears twice

Exception Handling

Definition: Follows alternative paths for error cases or special circumstances Characteristics:

• Includes exception-specific activities
• May be shorter (early termination) or longer (extra steps)
• Often branching to specialized handling

Example:

Application → Fraud Check → Investigation → Escalation → Final Decision
Cases: 12 (5%)
Avg Duration: 15.2 days
Special: Fraud investigation path

Using the Variants View

The variants page in Flow Myna provides a comprehensive overview:

Variant List

Default View: Sorted by frequency (most common first)

Each variant shows:

• Sequence of activities (visual or text)
• Case count and percentage
• Average duration
• Conformance type
• Performance metrics

Sorting Options:

• By frequency (most common first)
• By duration (slowest first)
• By conformance type
• Alphabetically

Variant Details

Click any variant to see:

• Complete activity sequence
• Step-by-step timing
• Case list (actual cases following this variant)
• Duration distribution
• Performance compared to other variants

Example Detail View:

Variant: Standard Approval Path
Cases: 175 (70%)
Conformance: Happy Path

Sequence:
1. Application Submitted (0h start)
2. Credit Check (+2.5h)
3. Risk Assessment (+28h)
4. Manager Approval (+48h)
5. Loan Disbursed (+72h)

Total Duration:
Mean: 4.2 days
Median: 3.9 days
P95: 7.1 days

Visual Representation

Variants can be displayed as:

• Text list: Activity1 → Activity2 → Activity3
• Timeline: Horizontal bars showing sequence and timing
• Compact notation: A → B → C → D

Filtering by Variant

One of the most powerful features: filter your process map to show only specific variants.

How It Works

1. Select one or more variants from the list
2. Click "Apply Filter" or "Show on Map"
3. Process map updates to show only those variants
4. Edges and nodes adjust to reflect filtered data

Use Cases

Compare Variants Select two variants to see differences:

• Standard path vs. Expedited path
• Happy path vs. Rework path
• High-value vs. Low-value processing

Investigate Problems Focus on problematic variants:

• Select all rework variants
• View only slow variants (> P95 duration)
• Isolate anomalies

Validate Improvements

• Filter to "before improvement" period
• Compare with "after improvement" period
• Measure impact on specific variants

Example: Comparing Standard vs. Expedited

Select both variants, process map shows:

• Shared steps in standard color
• Steps unique to standard path in one color
• Steps unique to expedited path in another color
• Clear visual of where paths diverge

Insight: Expedited path skips "Risk Assessment" and goes straight to "Quick Assessment", saving 1.5 days.

Common Variant Patterns

Recognizing these patterns helps you understand your process:

Linear Progression

Simple, straightforward path:

A → B → C → D → E
(No branches, no loops)

Meaning: Predictable, well-defined process

Early Termination

Process ends before reaching final step:

A → B → C → End
vs.
A → B → C → D → E → End

Meaning: Rejection, cancellation, or early completion

Skip Patterns

Some steps are bypassed:

A → B → D → E  (skips C)

Meaning: Conditional logic, business rules, automation

Rework Loops

Activities repeat:

A → B → C → B → C → D
(B and C repeat)

Meaning: Corrections, rejections, iterative refinement

Parallel Execution

Variants where different steps happen in different orders:

Variant 1: A → B → C → D
Variant 2: A → C → B → D

Meaning: Independent activities, flexible ordering

Escalation Paths

Extra steps for special cases:

Standard: A → B → C
Escalated: A → B → Manager Review → Director Review → C

Meaning: Approval hierarchies, risk management

Variant Analysis Best Practices

Start with the Top 5

Focus on the most frequent variants first:

1. Identify your top 5 variants
2. Understand why they differ
3. Measure their performance
4. Compare durations and outcomes

Look for Unexpected Patterns

Question variants that seem wrong:

• Steps in unexpected order
• Critical steps being skipped
• Excessive rework
• Inconsistent with policy

Measure Variant Stability

Track how variant distribution changes:

• Are new variants appearing?
• Is the happy path becoming less common?
• Are rework variants increasing?

Compare Cohorts

Analyze variants across different groups:

• High-value vs. low-value cases
• Different time periods
• Different regions or teams
• Approved vs. rejected cases

Using Variants for Improvement

Identify Standardization Opportunities

Question: Why do we have 15 variants when we designed one process?

Action:

• Analyze what causes variation
• Standardize where appropriate
• Document legitimate exceptions

Reduce Rework

Question: Why do 15% of cases have rework loops?

Action:

• Identify common rework triggers
• Improve upstream quality
• Add validation checks earlier

Optimize Common Paths

Question: Can we make the happy path faster?

Action:

• Focus on the 70% standard variant
• Remove unnecessary steps
• Automate manual activities

Handle Exceptions Better

Question: Are exception paths slower than they need to be?

Action:

• Design efficient exception handling
• Provide clear escalation paths
• Don't slow down happy path to handle exceptions

Next Steps

Now that you understand variants, continue your process exploration:

Analyze Your Process Map

• Process Map - See how variants appear in the flow diagram

Get AI Insights

• AI-Generated Insights - Let AI analyze your variants
• AI Co-Pilot - Ask "Why do we have so many variants?"

Advanced Analysis

• Filters & Exploration - Drill down into specific variants
• Event Groups - Simplify variant views