Skip to main content

Flows: Discover User Paths

Flows identifies the most frequent paths users take to or from any event. Use Flows to understand how users sequentially perform events in your product and analyze drop-offs or unexpected behavior. Flows Overview

Use Cases

Flows helps you answer path-based questions:

User Journey Discovery

  • What did users do immediately after signing up?
  • What are the most common sets of actions after opening the app?
  • What actions lead up to creating a channel?
  • How do users navigate between account creation and first message?

Drop-Off Analysis

  • Where do users exit the onboarding flow?
  • What paths lead to checkout abandonment?
  • What do users do instead of the desired action?

Feature Discovery

  • How do users discover key features?
  • What leads to feature activation?
  • What paths result in the highest engagement?

Quick Start

Let’s build a Flows report to answer:
What are the top paths leading Chrome users from Landing Page to Sign Up?

Step 1: Select Anchor Events

Anchor events are key moments you want to explore. Select “Landing Page” and “Sign Up”. Flows Steps

Step 2: Choose Steps Before/After

Decide how many steps to show:
  • Before the first anchor event - What led users here?
  • Between anchor events - What happens in the middle?
  • After the last anchor event - What happens next?
For this example, show 3 steps after “Landing Page”.

Step 3: Add Filters

Filter to Chrome users by adding a “Browser” filter where Browser = “Chrome”. Flows Filters

Step 4: Expand Events (Optional)

Break down any step by a property to see how flows differ:
  • Expand “Check Out” by “City”
  • See how flows vary by geographic location
  • Compare paths across different segments

Step 5: Apply Breakdowns

Breakdowns show parallel flows for different segments, helping you compare entire user journeys:
  • Compare flows for different cohorts
  • Understand segment-specific paths
  • Identify where paths diverge

Step 6: Analyze Results

Flows offers two powerful visualizations: User Flows (Sankey Diagram) User Flows
  • Width represents user volume
  • Follow paths visually
  • Click nodes to highlight top paths
  • Hover to see exact numbers
Top Paths (List View) Top Paths
  • Shows 50 most common sequences
  • Easy to compare exact paths
  • Clear conversion percentages
  • Step-by-step drop-off rates

Understanding the Sankey Chart

Anchor Events

These are the key events you selected to explore, labeled A, B, C: Anchor Events
  • A: Exit Tutorial
  • B: Experiment Started
  • Shows events performed between them

Intermediate Events

Mixpanel automatically shows the most common events performed before or after your anchor events: Intermediate Events
  • Regular events (tracked events)
  • Custom events (combined events)
  • Icon indicates event type

Other Events

By default, the top 3 events at each step are shown. Less common events are grouped as “Other Events”: Other Events Right-click to expand and view additional events.

Drop-Off

Users who didn’t perform any subsequent events in the selected time period: Drop-off
  • No further activity in time range
  • Didn’t complete next anchor event
  • Indicates exit points

Key Features

Adding More Steps

Expand your analysis dynamically: In the Visualization:
  • Click ”+” on the right to add steps after events
  • Click ”+” on the left to add steps before events
Add Steps Right Add Steps Left In the Query Builder: Adjust step counts in bulk: Add Steps Query

Multiple Anchor Events

Add multiple anchors to examine flows between specific events: Multiple Anchors Between Anchors:
  • ”+” left of breakpoint: Events after first anchor
  • ”+” right of breakpoint: Events before second anchor

Viewing More Events

Expand “Other Events” to see additional paths: Expand Events
  • Right-click “Other events”
  • Select “Add a row of events”
  • Or adjust in advanced settings

Conversion Criteria

Counting Method

How user activity is counted: Counting Method Uniques
  • One entry per user
  • First occurrence only
  • Best for understanding unique user paths
Totals
  • Multiple entries per user
  • Users can restart the flow
  • Best for measuring all path occurrences
Sessions
  • All steps in same session
  • Users can restart in new sessions
  • Best for same-session workflows

Conversion Window

Maximum time to complete the flow from first step: Conversion Window
  • Default: 30 days
  • Maximum: 366 days
  • Adjust based on user behavior

Hold Property Constant

Require same property value across all steps: Hold Property Example:
  • Hold “Item Name” constant
  • Tracks users interacting with same item
  • Filters out users who switch items
You can only hold properties that exist on all anchor events.

Advanced Features

Top Paths View

Switch to Top Paths for a list-based view: Top Paths View Shows:
  • 50 most common event sequences
  • Total users reaching each step
  • Conversion rates between steps
  • Overall completion percentage
Top Paths Details

Expand Event by Property

Break down a specific event by property: Expand by Property Methods:
  1. Click Expand button, select event and property
  2. Right-click event in chart, select “Expand by property”
Use Cases:
  • How does purchase type change the flow?
  • Do users from different cities follow different paths?
  • How does device type affect navigation?

Breakdown by Cohorts

Compare flows across different user segments: Cohort Breakdown Select cohorts to see color-coded parallel flows: Cohort Comparison Example: Compare “Android Users” vs “iPhone Users”
  • Different colors for each cohort
  • See path differences
  • Identify segment-specific behaviors
  • Click nodes for detailed breakdowns
Cohort Details

Hide Events

Simplify flows by hiding irrelevant events: Hide Events Methods:
  1. Click Hide Events button, select events to hide
  2. Right-click event in chart, select “Hide”
Use Cases:
  • Remove noisy tracking events
  • Focus on important actions
  • Simplify complex flows

Custom Events in Flows

As Anchor Events

Use custom events (groups of events) as start/end points:
  • Combines multiple events into one
  • Simplifies analysis
  • Shows either underlying event

As Intermediate Events

Mixpanel automatically shows common custom events: Custom Events Criteria:
  • Shared with all project users
  • Top 50 most-queried custom events
  • Non-overlapping definitions
Expand Custom Events: View underlying events when needed: Expand Custom Disable Custom Events: Show only regular events: Disable Custom

Exclusion Steps

Exclude users who performed certain events: Example:
  • Explore second most popular path
  • Remove users who saw errors
  • Focus on successful flows

Real-World Examples

Example 1: Onboarding Exploration

Query:
  • Anchor: Sign Up Completed
  • Show 5 steps after
  • Filter: Country = “United States”
  • Counting: Uniques
Insights:
  • 40% go to Profile Setup
  • 25% go directly to Dashboard
  • 20% explore Features page
  • 15% drop off
  • Action: Test prompting profile setup

Example 2: Purchase Path Analysis

Query:
  • Anchor A: Product Viewed
  • Anchor B: Purchase Completed
  • Show 3 steps between
  • Breakdown: Device Type
  • Hold “Product ID” constant
Insights:
  • Desktop: View → Compare → Add to Cart → Purchase (12%)
  • Mobile: View → Add to Cart → Purchase (8%)
  • Mobile users skip comparison step
  • Action: Simplify mobile comparison feature

Example 3: Feature Discovery

Query:
  • Anchor: App Opened
  • Show 4 steps after
  • Breakdown: User Cohort (New vs Returning)
  • Visualization: Top Paths
Insights:
  • New users: Tutorial (65%), Skip Tutorial (35%)
  • Returning users: Feature A (45%), Dashboard (40%)
  • Different onboarding paths
  • Action: Personalize based on user type

Example 4: Drop-Off Investigation

Query:
  • Anchor A: Checkout Started
  • Anchor B: Purchase Completed
  • Show 2 steps between
  • Exclude: Purchase Completed
  • Filter: Amount > $100
Insights:
  • 45% drop at payment info entry
  • Most go to FAQ page before dropping
  • Top question: “Is my data secure?”
  • Action: Add security badges to checkout

Tips for Effective Flow Analysis

Start with Key Events: Choose anchor events that represent important milestones or questions you want to answer.
Use Both Views: Sankey diagrams show overall patterns; Top Paths shows specific sequences. Use both for complete understanding.
Hide Noise: Hide irrelevant events (like tracking pings) to focus on meaningful user actions.
Compare Segments: Always break down by key segments (device, cohort, region) to find segment-specific insights.
Investigate Drop-Offs: Large drop-off percentages indicate friction. Click to view those users and understand why they left.
Hold Properties Constant: When tracking product-specific journeys, hold Product ID constant to avoid mixing different product flows.

Common Questions

Q: Why don’t I see my event in the flow? A: The event might not be in the top 3 at that step. Right-click “Other Events” and add more rows. Q: Can I see all possible paths? A: Flows shows the most common paths. Use filters and expansion to drill into specific paths of interest. Q: How is drop-off calculated? A: Drop-off shows users who didn’t perform any tracked events in the next step during the selected time period. Q: What’s the difference between Flows and Funnels? A: Funnels measure conversion through a specific path you define. Flows discovers what paths users actually take.

Next Steps

  • Learn about Funnels for conversion analysis
  • Create Cohorts from flow segments
  • Build Boards to monitor key flows
  • Explore Retention to track engagement

Build docs developers (and LLMs) love