User Path Analysis

Analyze forward and reverse user paths in MoEngage to see how users navigate before and after key events on your website or mobile app.

Step 1: Choose Event

An event refers to any action that a user completes on your website or app that is tracked. It’s important to understand how users navigate after completing their first event and to know the path they take to reach a specific event. You have two options to analyze:

• Forward user paths: This option displays all user paths that begin with the selected event. To use this feature, select the Path type as Starting with and select the desired Event. For example, you can analyze the user paths that start with the event App/Site Opened.
• Reverse user paths: This option shows all user paths that lead up to the selected event. To use this feature, select the Path type as Ending with and select the desired Event*.* For example, you can analyze the user paths that end with the event Checkout Completed.

The event can be further filtered. To do so, click + Attributes and select event attributes for specific user actions, such as Platform, App Version, and so on.

Click + Attributes again and combine multiple attributes using AND/OR for each event. For example, use the OR filter to combine data across two platforms. If you want to get data for a specific platform and look at the visits to a specific website, use the AND filter.

Step 2: Filter Users

By default, you can obtain the user paths for all users. However, you can also query for a group of users filtered using User property, User activity, User affinity, or Custom segments. Filtering users is similar to creating a segment. For more information, refer to Create Segment.

Step 3: User Paths Options

User Path Window

Select the desired time window to analyze user paths within a specific duration. This option considers the events that occur within the specified time after the starting event or before the ending event.

The window is set to four hours by default. You can choose units of Days, Hours, or Minutes.

For example, to analyze user paths that occur after the App/Site Opened event within a span of 10 hours, you can set the user path window accordingly.

User Path Steps

By default, the user paths chart shows the first five steps taken by the users after the selected event. For some user events, you might want to analyze the path for a specific number of steps. You can have up to 15 steps in user paths.

For example, analyze the first three actions of the user after they receive a push notification by setting the User path steps to three.

Date Range

After defining the user path window and user path steps, select the time frame for which you want to perform the user path analysis. You can choose from:

• Today
• Yesterday
• This Month
• Last Month
• Since
• Before
• Custom range

By default, the date range selected is Yesterday. User paths cannot be analyzed for more than 30 days at a time. After selecting the date range, click APPLY. Your chart is now generated.

Step 4: Chart

User paths are illustrated using a Sankey Chart analysis, which displays paths and steps flowing from left to right. Each Node represents a specific event, while a vertical arrangement of nodes signifies a step. The area or cord connecting two nodes indicates the path taken.

You can analyze user paths by examining consecutive nodes and paths. The forward user path begins at an event and then branches out into multiple paths taken by different users following that initial event. In contrast, reverse user paths involve the merging of different paths into one event.

Chart for the Forward User Paths

When you hover over a node, the total number of users for that event is shown, as well as a percentage. You can also see the percentage of users who went to the next step and the percentage of users who did nothing after the current step and dropped off. In this scenario, the end nodes do not display any information about the drop-off or the next step because no information is available.

Drop off node: A dedicated node showing the total number of users who dropped off after the previous step, shown as a percentage of the total users at the starting node.

Paths represent the number of times users went from one node to another, along with the origin and destination node names, and the percentage for the origin node.

Chart for Reverse User Paths

When you hover over a node, you see the total number of users at that event, along with a percentage relative to the total users who reached the final (anchor) event you selected. For each node, you can also see:

• From previous step: The percentage of users who reached this node by completing a tracked event at the step before it.
• Entry: The percentage of users who reached this node with no tracked event before it in the path window. These users enter the path directly at this step.

The leftmost (start) nodes have no earlier step to reference, so they don’t show a “from previous step” percentage. Every user at a start node counts as an entry.

Entry node: A dedicated node showing the total number of users who entered the path at that step, meaning no prior event was tracked for them within the user path window. This is shown as a percentage of the total users who reached the next event. It’s the reverse-path equivalent of the Drop off node in forward paths: Drop off counts users exiting a path early, Entry counts users joining a path partway through without an earlier tracked step.

Paths represent the number of users who moved from one node to another, along with the origin and destination node names, and that count as a percentage of the origin node’s total.

Chart Settings

Repeated Events

Repeated events are a series of the same events performed multiple times by users in succession. Click Hide to merge the sequence into one event, and click Show to view the entire sequence.

Exclude Events

To hide certain events from appearing in your User Paths chart, select the desired events from the Excluded events list and click Apply Filters.

Alternatively, you can add the events to be excluded directly to the chart. To do so, perform the following steps:

1. On the user paths chart, click a desired node. The Perform action dialog box appears.
2. Under Actions, select Exclude this event.
3. Click Apply.

Resolution

Resolution sets a minimum percentage threshold for a node to appear on its own in the chart. At any given step, events that fall below the threshold are grouped into a single node labeled Others instead of appearing individually. Raise the resolution to simplify the chart to major paths only; lower it to see more detail.

Split By

Split By breaks a node into separate branches based on an event or user property you choose. For example, splitting the App Opened event by Platform creates separate branches for iOS and Android. When applied to an event, every node where that event appears is split by the selected attribute.

You can select a maximum of five split by attributes for any step.

Alternatively, you can add the split by attributes directly on the nodes. To do so, perform the following steps:

1. On the user paths chart, click a desired node. The Perform action dialog box appears.

2. Under Actions, select Split by.

3. From the Select attributes list, select the desired split by attributes. You can split an event by up to five attributes.

4. Click Apply.

   

Zoom Chart

While analyzing user paths with a higher number of steps, you can use the Zoom chart functionality to zoom in on a specific step or node and view the details clearly. Similarly, zoom out shall zoom out the chart, and reset zoom shall reset the zoom level of the chart.

Add and Remove Steps

Apart from the Zoom functionality, you can click Add Step or Remove Step to control the number of visible steps in the chart. These buttons are available on the right side of the chart in the forward user path and the left side of the chart in the reverse user path.

Why Results Can Vary Between Runs

Running the same User Path analysis twice — same events, path window, steps, and date range — can return slightly different numbers each time.

This happens when a user has multiple events ingested with an identical timestamp. MoEngage Analytics reads event data from the warehouse as it was ingested, and events that share the same timestamp have no deterministic relative order. The path computation can resolve that order differently on each run.

This is expected behavior rather than a defect or a sign of missing data. The variance comes from the event data itself, not from the analysis.

To reduce it, space out the timestamps of events that your app or server sends in quick succession, so that each event carries a distinct time. For how same-timestamp events are drawn on the chart, refer to What causes unexpected event sequences in user paths?.

Download

After you view the chart, you might need to share it with your team. Click Download Chart in the upper-right corner to download it. The chart will be saved in PNG format.

Save to My Studio

Your analysis is now ready. To access this later, you can save the analysis.

Perform the following steps to save your analysis:

1. Click the icon in the upper-right corner.

2. Click Save to My Studio.

3. Enter a name and description for your analysis.

4. Click Save.

   

You can access the analysis by navigating to Analyze > My Studio.

Next Steps

• MoEngage Dashboards
• My Studio