The WebEngage Knowledge Base Developer Hub

Welcome to the WebEngage Knowledge Base developer hub. You'll find comprehensive guides and documentation to help you start working with WebEngage Knowledge Base as quickly as possible, as well as support if you get stuck. Let's jump right in!

Get Started    Discussions

Analyzing Paths

A step-by-step guide to analyzing how users interact with your platforms & where they drop-off

So without much ado, let's dig in!

How It Works

🚧

Must Read

Please ensure that you have a robust understanding of Events & Event Attributes before proceeding as these data points track user behavior and form the basis of Path analysis.

  • Insights gained through the Path Lab are as good as the behavioral data you track for your apps and websites. Thus, we highly recommend that you:

    • Step 1: Map the various user lifecycles of your product and highlight crucial interactions at each stage.

    • Step 2: Track all relevant user actions as Custom Events, and attach unique Custom Event Attributes to each. (Start Here) Doing so will enable you to:

      • Visualize the various flows taken by users in your app and website to achieve their end-goal like, discover products/content/courses/services, make a purchase, evaluate services/subscription models/flights, and so on.

      • Analyze platform interactions of a niche set of users that purchase a specific type of subscription, place high-value orders, consume content of a specific genre, and so on.

  • The various flows taken by users to interact with your app and website are plotted as a Sankey diagram, where each Event performed by them is depicted as a Step. (detailed read).

    • These sequential Steps are plotted as per a time frame or Lookahead Window defined by you and are not limited to a user's actions in an on-going session.

Now, let's show you how you can leverage this tool to get granular behavioral insights!

Configure a View

The query tab enables you to drill down into several variations of the Paths users choose to interact with your app and website.

Click to enlarge

Here's how you can build a custom Path view for analysing specific user behavior:

Step 1: Select a Time Frame

Click to enlarge

While the default time frame of analysis is set to 7 days, using the date range filter placed on the top right, you can choose to analyze behavioral trends for the following durations:

📘

  • Today

  • Yesterday

  • Lasy 7 days

  • Lasy 30 days

  • Lasy 90 days

  • Custom dates

Step 2: Specify Scope of Analysis (Show Path)

This is the most crucial step in building a Path view as the anchor Event and context selected here form the primary reference point of your analysis.

Click to enlarge

Step 2.1.: Specify the Context

As shown above:

  • Select Starts With to analyze the series of Events users perform after the anchor Event on your app/website.

  • Select Ends With to analyze the series of Events users performed before the anchor Event on your app/website.

Step 2.2.: Select Anchor Event

You can think of the anchor Event as the starting point, based on which, various flows will be plotted, highlighting the most common series of Events performed by users before or after performing it.

For example, in the above visual we've specified the Custom Event, Cart - Viewed as the anchor Event. As a result, we're able to analyze the various actions user perform after it.

Let's go over a short use-case to show you how it works:

👍

Use-case: What do users do after they initiate checkout?

Let's take the example of an e-commerce app. While their userbase continues to grow at a steady pace, the purchase rate is lagging behind.

A quick funnel analysis indicated that most users drop-off the Checkout Flow after initiating the transaction. This prompted its Product Manager to dig deeper into what users do once deviate from the checkout screen.

Pre-requisite:

  • Each time a user clicks the checkout button in their cart, it's tracked as the Custom Event, Cart - Checkout Started.

  • Each time a user completes a transaction, it's tracked as the Custom Event, Checkout Completed.

Thus, to analyse the above use-case, we will define the following scope:

Show Paths that Start With the Event, Cart - Checkout Started.

This means: You will be able to analyze the most common series of actions taken by users after they perform the Anchor Event, Cart - Checkout Started.

Similarly, you can choose to trace back user actions leading up to purchase by specifying the following scope:

Show Paths that End With the Event, Checkout Completed.

👍

Comparing the Path views mentioned above will help the e-commerce app's Product Manager identify various factors that causes users to deviate from the intended Checkout Flow.

Step 2.3. (Optional): Add Attribute Filters to Anchor Event

You can further narrow down the scope of occurrence the anchor Event by adding attribute filters to it. As shown below, this can be done by clicking the filter icon placed next to the field, Event.

Click to enlarge

For example, in the above visual we have defined the scope of occurrence of the anchor Event, Cart - Viewed by adding the attribute filter, Cart Value greater than 1,000. Thus, Paths will be plotted for only those users who have products worth $1,000 or more lying in their cart.

📘

The following event attributes can be applied as filters:

  • Custom Attributes (all the custom attributes attached to the custom event selected as the anchor event)

  • Time (Event Time)

  • Location (City, Country)

  • Technology (Browser Name, OS Name, Device Manufacturer, Device Model, Carrier, App Version, App ID, Platform, SDK Version)

  • UTM (Channel, Campaign Name, Campaign Source, Campaign Medium)

  • Screen (Page URL, Screen Name)

  • Engagement (Campaign ID, Journey ID)

Adding Multiple Attribute Filters to the Anchor Event

Depending on your analytical needs, you can choose to analyze platform interactions of a niche set of users who perform the anchor Event in the context of All/Either attribute filters. As shown below, click Add Filter and club multiple event attributes by the AND/OR logic.

Click to enlarge

For example, in the above visual we have defined the scope of analysis as: Show Paths starting with the Event, Cart - Viewed for all users who's Cart Value is greater than or equal to $1,000 OR users who have added 10 or more products to their cart (Cart Size). Thus, users who don't match either of these criteria will be excluded from the analysis!

🚧

Implications of using AND/OR to combine attributes

AND: Clubbing filters by the AND logic helps you narrow down the scope of occurrence of the anchor Event. This means that paths will be plotted for only those users who perform the Event in the context of all the event attributes.

OR: Clubbing filters by the OR logic helps you widen the scope of occurrence of the anchor Event. This means that paths will be plotted for all users who perform the Event in the context of any of the specified event attributes.

Step 3: Specify Lookahead Window

You can think of the lookahead window as an activity time frame that helps you determine exactly how long before/after performing the anchor Event would you like to analyze users' Paths.

Thus, the flows depicted in the Path visualizer are based on the sequential actions users take within the time frame defined here. This is independent of whether the actions were performed in an ongoing session or multiple sessions.

The Lookahead Window can be specified in Minutes or Hours. For example, we've chosen to analyze the Events performed by users up to 48 hours before they perform the anchor Event, Subscription Purchased in the visual below.

Click to enlarge

Step 4 (Optional): Split By

You can choose to add a second dimension to your analysis by splitting the Anchor Event. In doing so, you will be able to compare the most common flows taken by different groups of users that performed the Anchor Event in varying contexts.

This comes in handy when you're looking to optimize the platfrom experience of a niche set of users who:

  • Use a specific version of a browser or your app
  • Interact with a specific app screen/ web landing page
  • Engage with a specific genre of content/ product category/ educational subject - and so on.

Click to enlarge

For example, in the above visual, we've chosen to split the view by Country. This means:

  • All users who belong to a Country will be grouped under a common starting point.
  • The flows depicted on the visualizer indicate the most common series of actions performed by users in each Country, before/after the Anchor Event, within the specified Lookahead Window.

📘

The following event attributes can be selected here:

  • Location (City, Country)

  • Technology (Browser Name, OS Name, Device Manufacturer, Device Model, Carrier, App Version, App ID, Platform, SDK Version)

  • UTM (Channel, Campaign Name, Campaign Source, Campaign Medium)

  • Screen (Page URL, Screen Name)

  • Engagement (Campaign ID, Journey ID)

  • Custom (all the custom attributes attached to the custom event selected as the anchor event)

Step 5 (Optional): Exclude Events

Using this field, you can choose to omit a maximum of 10 Events from the Path. It enables you to declutter noisy data and focus only on what matters. As shown above. you can select from a list of all the Events tracked for your account.

Click to enlarge

For example, in the above visual, we've removed generic System Events like Session Started, User Login and User Logout. This helps us focus on crucial Events performed by users as they experience the Checkout Flow.

Step 6 (Optional): Include Events

Using this field, you can choose to analyze the correlation between a maximum of 10 Events performed by users in your app and website. Let's go over a short use-case to help you understand this better:

👍

Identifying Paths Taken by Users to Evaluate Hotel Stay After Booking Flight Tickets

Let's take the example of a travel booking app that enables users to book Flights and Hotels. While they're experiencing steady growth in individual bookings, their cross service bookings remain low. (aka users that book Flights + Hotel Stay)

Marketers of the app took it upon themselves to drive cross-sales. They started their analysis with the question - How many users search for hotels, select a location, view hotel details, and book a hotel stay after purchasing flight tickets?

Platform interactions mentioned above are tracked as the following Custom Events:

  • Search for hotels = Hotel Search Clicked

  • Select location = Hotel Location Selected

  • View hotel details = Hotel Details Viewed

  • Book stay = Hotel Booking Confirmed

  • Purchasing flight tickets = Flight Booking Confirmed

The idea was to identify the most common Paths taken by users to perform specific Events sequentially in each case listed above. evaluate another service. Marketers can then leverage various moments along these paths to engage users with personalized messages, motivating them to convert.

Thus, they configured a view by adding the following Events to the field, Include Events:

Click to enlarge

Similarly, you can click the dropdown and select from a list of all the Events tracked for your account to analyze the various sequences in which users perform a specific set of actions.

Here's what their analysis revealed:

Click to enlarge

Step 7 (Optional): Collapse Repeated Events

There are several actions that users repeatedly perform while interacting with your app/website. Enabling this field helps you eliminate redundancy by minimizing multiple occurrences of Events to one.

Click to enlarge

For example, most e-commerce app users would Browse through several products and sections before making a purchase. This means, their Path could look something like this:👇

👍

Anchor Event > Step 2: Browse > Step 3: Browse > Step 4: Browse > Step 5: Sort Applied > Step 6: Browse > Step 7: Browse > Step 8: Product Viewed

Thus, you can gain sharper insights by hiding repeated events that are inconsequential to your analysis. In doing so, the flow will indicate the Event only the first time it occurs, the subsequent Step will depict the next (new) Event users perform, like this:👇

👍

Anchor Event > Step 2: Browse > Step 3: Sort Applied > Step 4: Browse > Step 5: Product Viewed

Neat, isn't it?

Next, click Show Paths to begin analysis!

Analyze Paths

Once you execute your use-case through the Query Tab, the most common Paths taken by users within the specified Lookahead Window, before/after the Anchor Event are plotted below it.

You can analyze these platform interactions in 2 formats, through the following sections:

  • Overview: Includes a Paths visualizer or a Sankey diagram in which width of the lines is proportional to the flow quantity, aka. the number of users that opted to experience a flow.

  • Top Paths: Summarises the top 10 Paths your users experienced before/after they performed the Anchor Event on your app/website along with a step-wise conversion rate.

Let's get you acquainted with each:

Overview

The visualizer comes equipped with several features that enable you to:

1. Steps

  • Events performed by users before/after the Anchor Event in your app/website are indicated as subsequent Steps in the Path visualizer.

  • By default, the visualizer plots Paths containing a maximum of 4 Steps where:

    • Step 1 is the Anchor Event selected by you
    • Step 2 is the Event performed immediately before/after it, and so on.
  • Each Step lists:

    • A maximum of 5 Events: You can interpret this as the top 5 Events users choose to perform in a specific moment in their lifecycle. Each Event indicates the percentage of users that performed it.

    • Drop-offs: Indicates the share of users that exited your platform in that specific moment. This is crucial insight that can be leveraged to:

      • Prolong platform engagement.
      • Drive users towards a specific goal.
      • Correct the existing UI flow of your platform to prevent drop-offs - and much more!
    • Others: Several different Events performed by a small share of users at a Step is indicated by the blanket term, Others. These statistically insignificant interactions have been hidden to ensure that you are presented crisp insights.

Adding/ Removing a Step

Using the operator placed on the top right, you can choose to:

  • Reduce the number of Steps plotted for all the Paths. (click -)

  • Add upto a maximum of 12 Steps to widen the scope of analysis. (click +)

Removing an Event

There are several instances where the Paths visualizer includes Events that are inconsequential to the problem statement you're analysing. This is why, we've made it extremely easy for you to exclude such actions from the scope of your analysis.

For example, in the visual below, we're analyzing the Credit Card Risk Calculator flow for an Insure-tech platform. While the flow is intended to look like this:

👍

Click Calculate Risk Button --> Start Quiz --> Answer 7 Questions --> End Quiz (where risk % is indicated for each user) --> Evaluate Credit Card Protection Plans --> Purchase

You can clearly see a number of deviations and Drop-offs across the various Paths chosen by users after they perform the Anchor Event, Calculate Risk. (Only 5% actually make it to the purchase section!)

Click to enlarge

As shown above, here's how we refined the data plotted across all the Paths by excluding several Events:

Step 1: Hover over an Event you'd like to exclude from visualizer.

Step 2: Using your trackpad/mouse, right click on the Event and select the option, Remove Event.

Analyzing Flow of Platform Interaction From a Specific Event

Click to enlarge

Step 1: Hover over an Event you'd like to analysis in further detail and click on it.

  • In doing so, all the Paths connecting the Event, across all the Steps will be highlighted. This makes it easier for you to visualize the exact route users experienced before and after performing the Event.

Step 2: Using your trackpad/mouse, right click on the Event and select the option, Remove Highlighting to deselect it.

2. Paths

Click to enlarge

Each Event listed under a Step in the visualizer is connected by several Paths that indicate the exact route users took in interact with your app/website. The thickness of each Path is directly proportional to the number of users that experienced it.

Let's quickly walk you through all the details indicated for each Path:

First Step: Indicates the Anchor Event specified by you and the total number of times users have performed it within the selected time frame.

  • The Anchor Event is referred to as the Last Step in the visualizer only if the scope of analysis has been defined as - Show Paths that End with

Selected Path: Indicates the total number of users that experienced the path selected by you within the selected time frame and Lookahead Window.

Conversion Rate: Indicates the share of users that experienced the Path selected by you, out of all the users that performed the Anchor Event.

We hope this helps you gain actionable insights into how users engage with your platform and identify drop-off areas that can be optimized by delivering personalized experiences. Please feel free to drop in a few lines at [email protected] if you have any further queries. We're always just an email away!

Updated about a month ago

Analyzing Paths


A step-by-step guide to analyzing how users interact with your platforms & where they drop-off

Suggested Edits are limited on API Reference Pages

You can only suggest edits to Markdown body content, but not to the API spec.