Process Map

The Process Map is your visual window into how your process actually works. It's an interactive, node-based diagram that shows all activities, their connections, and detailed performance metrics. This is the heart of process mining.

Understanding the Process Map

The process map automatically visualizes your entire process flow by analyzing your event data. Instead of manually drawing flowcharts based on assumptions, you see the real process as it actually happens.

What You See

The process map shows:

• Nodes: Activities/events that occur (represented as boxes)
• Edges: Transitions between activities (represented as arrows)
• Metrics: Frequencies, durations, and performance statistics
• Paths: All the ways cases flow through your process

Example: Loan Application Process

In a typical loan process map, you might see:

• Application Submitted → Credit Check → Risk Assessment → Approval → Disbursement

But the map also reveals:

• Exception paths (rejection routes)
• Rework loops (resubmissions)
• Parallel activities (concurrent processing)
• Bottlenecks (where things slow down)

Reading Nodes and Edges

Understanding what nodes and edges represent:

Nodes (Activities)

Each node represents an event type - an activity that occurs in your process.

What nodes show:

• Activity name (e.g., "Credit Check Complete")
• Total cases that experienced this activity
• Percentage of all cases
• Position in the process flow

Example node:

Each node displays two key numbers:

• Green number (left): The count of objects that have entered and passed through this event
• Grey number (right): The count of objects currently waiting at this event—they reached this step but haven't moved to the next event yet (the process is still in progress for these cases)

This distinction helps you understand not just historical flow, but also where work is currently sitting in your process.

Node colors:

• Standard: Normal flow activities
• Start: Process entry points (typically darker/highlighted)
• End: Process completion points
• Selected: Currently selected for detailed view (highlighted in emerald)

Edges (Transitions)

Each edge represents a transition - when cases move from one activity to another.

What edges show:

• Direction of flow (arrow)
• Number of cases that took this path
• Percentage of cases from source node
• Average duration between activities
• Visual thickness (thicker = more frequent)

Example edge:

Credit Check → Risk Assessment
175 cases (75% of credit checks)
Avg: 1.5 days, Median: 1.2 days

Edge interpretation:

• Thick edges: Common paths (most cases follow)
• Thin edges: Rare transitions (exceptions)
• Back-pointing edges: Rework loops (cases going backwards)
• Multiple outgoing edges: Decision points (cases split to different paths)

Duration Statistics

Edges display rich performance data to help you understand timing:

Key Metrics Explained

Mean (Average)

• Simple average of all durations
• Good for overall sense of timing
• Can be skewed by outliers

Median

• Middle value (50th percentile)
• Better representation of "typical" case
• Not affected by extreme outliers

P95 (95th Percentile)

• 95% of cases complete faster than this
• Identifies the slower cases
• Good for setting SLA targets

Min and Max

• Fastest and slowest observed cases
• Reveals range of variation
• Extremes may indicate issues or special handling

Standard Deviation

• Measure of variability
• High deviation = inconsistent timing
• Low deviation = predictable process

Example Duration Display

Application Submitted → Credit Check
─────────────────────────────────────
Cases: 250
Mean: 2.5 hours
Median: 1.8 hours
P95: 6.2 hours
Min: 15 minutes
Max: 2.3 days
Std Dev: 1.9 hours

What this tells you:

• Typical case (median): ~2 hours
• Some take much longer (max 2.3 days)
• High variability (std dev almost as large as mean)
• 5% of cases take over 6 hours
• Possible bottleneck worth investigating

Interactive Features

The process map is highly interactive, allowing you to explore in depth:

Zoom and Pan

Zoom Controls:

• Mouse wheel: Zoom in/out
• Zoom buttons: Click +/- controls
• Fit to view: Reset to see entire map

Pan Controls:

• Click and drag: Move the map around
• Helpful for large, complex process maps
• Navigate to specific areas of interest

Selecting Nodes

Click any node to:

• Highlight it in emerald green
• Show detailed statistics in info panel
• See all incoming and outgoing edges
• View cases that went through this activity

Node details include:

• Total cases
• Percentage of all cases
• Average time spent in this activity
• Frequency of occurrence

Selecting Edges

Click any edge to:

• Highlight the transition
• Show duration statistics
• See case counts and percentages
• View transition frequency

Edge details include:

• All duration metrics (mean, median, P95, etc.)
• Number of cases
• Percentage from source node
• Performance indicators

Full-Screen Mode

Toggle full-screen to:

• Maximize map visibility
• Remove sidebar distractions
• Focus on process exploration
• Useful for presentations or complex maps

Identifying Bottlenecks

The process map helps you quickly spot performance issues:

Visual Indicators

Color Coding Edges are colored on a spectrum from grey to red based on duration:

• Grey edges: Fast transitions (shorter duration)
• Red edges: Slow transitions (longer duration, potential bottlenecks)

The redder an edge appears, the longer cases typically take to move through that transition.

Thickness

• Thick edges: High volume (many cases)
• Thin edges: Low volume (few cases)

What to Look For

1. Long Duration Edges Red or highlighted edges with high mean/median times:

Risk Assessment → Manager Approval
Mean: 4.2 days ← BOTTLENECK
Median: 3.8 days
P95: 8.1 days

2. High Variability Large standard deviation indicates inconsistent processing:

Document Verification → Final Decision
Mean: 2.1 days
Std Dev: 3.5 days ← INCONSISTENT
(Some very fast, some very slow)

3. Rework Loops Back-pointing edges showing repeated work:

Final Decision → Document Resubmit → Review
(Cases looping back for corrections)
```text
**4. Split Points with Delays**
Decision points where one path is much slower:
```text
Risk Assessment → Auto Approval (fast, 2 hours)
Risk Assessment → Manual Review (slow, 3 days) ← BOTTLENECK

Example: Finding Your Biggest Bottleneck

In our loan example, the map reveals:

1. Risk Assessment → Manager Approval: 4.2 days average
2. Document Verification → Final Decision: 3.1 days average
3. Application → Credit Check: Only 2 hours average

Insight: Manager approval is the primary bottleneck, taking 2x longer than document verification and 50x longer than credit checks.

Using Filters

Filters let you drill down into specific scenarios:

Applying Filters

Filters narrow the process map to show only matching cases:

Example filters:

• "High-value loans only" (amount > $50,000)
• "Rejected applications" (ended at rejection node)
• "Cases over 14 days" (duration > 2 weeks)
• "Specific time period" (Q1 2024)

What changes:

• Node counts update (fewer cases)
• Edge frequencies adjust
• Duration statistics recalculate
• Some paths may disappear (if no matching cases)

Filter Impact

Before filter (all 250 loans):

Application → Credit Check: 250 cases, 2h avg
Credit Check → Approval: 175 cases, 1.5d avg
Credit Check → Rejection: 75 cases, 0.8d avg

After filter (high-value loans only, 80 cases):

Application → Credit Check: 80 cases, 2.1h avg
Credit Check → Approval: 65 cases, 2.3d avg ← Slower!
Credit Check → Rejection: 15 cases, 1.2d avg

Insight: High-value loans take longer to approve (2.3 days vs. 1.5 days overall)

Combining Multiple Filters

Filters combine using AND logic:

• Filter 1: "High-value loans"
• Filter 2: "Longer than 14 days"
• Result: High-value loans that took over 2 weeks

The process map updates in real-time as you add or remove filters.

Learn more: Filters & Exploration

Common Patterns to Recognize

Happy Path

The thickest edges forming a clear route:

Start → Step1 → Step2 → Step3 → End
(Most cases follow this standard route)

Exception Handling

Thin edges branching off to error handling:

Step2 → Exception Handler → Recovery → Step3
(Rare but important error paths)

Rework Loops

Edges pointing backwards:

Review → Rejected → Resubmit → Review
(Cases going in circles)

Parallel Paths

Multiple edges leaving a node:

Received → Path A (fast track)
        → Path B (standard)
        → Path C (complex review)

Skipped Steps

Missing edges you'd expect to see:

Expected: A → B → C
Actual: A → C (B was skipped)

Best Practices

First-Time Exploration

1. Start with full view: See the entire process
2. Identify start and end: Understand entry/exit points
3. Follow thick edges: Find the happy path
4. Look for red edges: Spot bottlenecks
5. Check rework loops: Find inefficiencies

Regular Monitoring

• Weekly review: Check for new bottlenecks
• Compare periods: This week vs. last week
• Monitor metrics: Are durations increasing?
• Track changes: Did process improvements help?

Presentation Tips

• Use full-screen mode: Maximum visibility
• Zoom to problem areas: Focus attention
• Use filters to compare: Show before/after
• Highlight specific paths: Tell the story clearly

Next Steps

Now that you understand the process map, explore other analysis tools:

Analyze Variants

• Process Variants - See different execution paths through your process

Get AI Insights

• AI-Generated Insights - Let AI identify opportunities
• AI Co-Pilot - Ask questions about your process

Advanced Exploration

• Filters & Exploration - Drill down with advanced filtering
• Event Groups - Simplify complex maps