Conversion and Attribution

Overview

When your campaigns are live, it's time to analyze their performance to understand which campaigns have performed well and which ones are yet to. Conversions are important metrics to track to understand how impactful your campaign is and whether you are able to achieve what you set out to do with your campaign.

Attribution refers to conversion attribution for a campaign and helps you understand how a conversion goal was fulfilled — whether the user viewed your campaign or clicked a link or notification and then fulfilled the goal set. For example, if the conversion goal for your campaign is an Add To Cart event, attribution helps you understand when and how the user added the product to the cart — was it after seeing your push notification, was it after clicking on the notification, or was it when they were already on the app looking for something.

You can use the following attribution models that you can use for tracking your campaign's performance in MoEngage: the All Interactions Model and the Last Interaction Model.

All Interactions Model

The All Interactions Model attributes conversions to campaigns based on the following attribution types:

Attribution Type Description
View-through conversion The total number of conversion goals executed by users who received your communication within the configured attribution window from the time of receiving the campaign.
Click-through conversion The number of conversion goals executed by users who clicked the notification within the configured attribution window from the time of click. Click-through attribution is click-based tracking. 
In-session conversion The number of conversion goals executed by users within 30 minutes after clicking your communication. The in-session attribution is based on In-session tracking. 

Attribution Window

The attribution window is the duration for which the conversion goal is tracked for users from the time they receive or click a communication from a campaign or Flow. For more information, refer to Attribution Window for Campaigns and Flows

How does conversion attribution work?

Conversion Attribution in Campaigns Conversion Attribution in Flows Conversion Attribution in Campaign Vs. Flows

The following examples describe how conversion attribution works in campaigns.

 

Example 1

You send the following campaigns to the users:

  • A Push campaign notifying the user of the latest offers (C1)
  • An In-app campaign that has banners for the Deal of the Day (C2)

The conversion goal for both campaigns is defined as the users adding an item on the Deal of the Day sale to their cart. Both campaigns have an attribution window of 20 hours.

Suppose you send the Push notification at 7 PM and the In-app notification is shown to the user around 8 AM the next day. 

If the user adds an item to the cart from the Deals page at Conversion Attribution
10 PM on Day 1 Push Campaign
9 AM on Day 2 Push Campaign and In-App Campaign
8 PM on Day 2 In-App Campaign

 

Example 2

You send the following campaigns to the users:

  • A Push campaign notifying the user of the latest offers (C1)
  • An Email campaign about upcoming offers (C2)
  • A WhatsApp campaign about lightning deals (C3)

The conversion goal for all these campaigns is defined as the users viewing the deals web page, and the campaigns have an attribution window of 36 hours, 20 hours, and 18 hours each.

You send the Push (C1) is sent at 11 PM on Day 1, the WhatsApp at 8 AM the next day (Day 2), and the Email campaign at 11 AM the next day (Day 2).

User views the product page at Conversion Attribution
6 AM on Day 2 Push Campaign
9 AM on Day 2 Push and WhatsApp Campaigns
9:30 AM on Day 3 Push Campaign

Thus, in the All Interactions Model, when there is an overlap of the attribution window between two campaigns having the same conversion goal, the conversion is attributed to both campaigns. To avoid this, you can time your campaigns with the same conversion goals such that their attribution windows do not overlap. For information about setting the attribution window for campaigns, refer to Set Attribution Window.

In some cases, you need to run multiple campaigns with the same conversion goals simultaneously. This can lead to conversions being attributed to multiple campaigns, and you may not get clear insights into which campaign is most relevant and which channel or flow gives you the best engagement and ROI. Having a robust conversion attribution model for campaigns is thus the need of the hour, and the Last Interaction Model for conversion attribution does that for you!

The Last Interaction Model helps you attribute conversions to the right campaign, thus helping you finetune your marketing strategies and leverage the insights gleaned towards improving customer engagement and ROI.

Last Interaction Model

The last interaction model is a conversion attribution model that considers the deepest and latest interaction of the user and attributes the conversion to the campaign with the deepest and latest interaction. Refer to Event Priorities for information about deepest interaction.

In the following scenario, a user receives a Push notification and executes the conversion goal within the configured attribution window.

  • It is a view-through conversion if the user does not click the notification and executes the conversion.
  • It is a click-through conversion if the user clicks the notification.

The latest interaction refers to the campaign with which the user interacted most recently. Interactions are ranked in depth in the following manner: Clicks > Views > Sent 

The Last Interaction Model attributes conversions to campaigns based on the following attribution types:

Attribution Type Description
Total conversions The sum of all conversions — click-based conversion and any other conversion such as view-based conversion or cross-platform attribution.
Click-through conversion The number of conversion goals executed by users who clicked the notification within the configured attribution window from the time of click. Click-through attribution is click-based tracking. 

How does the Last Interaction Model work?

The last interaction model attributes conversions in the following order:

  • deepest interaction
  • latest interaction

Thus, irrespective of the latest interaction, if there is a deeper interaction, the deeper interaction gets the attribution. For example, a click event trumps a view event. In the following example:

SMS Campaign C1 and Email Campaign C2 are live and have the same conversion goals. They are received by the same user. C1 and C2 have an attribution window of two days. John is the user who:

  • Clicks on the SMS from C1 on Day 1 at 2 PM.
  • Opens the Email from C2 on Day 2 at 10 AM.
  • Adds a product to his cart (conversion goal for both C1 and C2) on Day 2 at 10 PM.

In this case, the conversion is attributed to the SMS campaign C1 even though the Email Campaign C2 has the latest interaction because C1 has the deepest interaction.

Event Priorities for conversions

The following are the priorities assigned to various events that result in conversions:

  • Campaign clicks or the click event receives the highest priority because the user shows intent and acts on it as well by clicking on the communication received. Thus, if a user clicks a Push notification, SMS message, WhatsApp message, In-App message, On-site message, or a Card and achieves the conversion goal, that event receives the highest priority.
  • Campaign views (or impressions) or the view event is ranked below the click event because the user is aware of the campaign by seeing it. Thus, if a user receives a Push notification, opens an email,  reads a WhatsApp message, or has an In-app message, OSM, or Card shown to them, these are only views for the respective campaigns and would receive a priority lower than clicks.
  • Campaign sends, or the sent event, is ranked the lowest among the lot as the campaign is only sent to the user, and they may not be aware of it or act on it. Thus, if a Push notification, SMS, email, Whatsapp message, or Card is sent to the user, these are sends and receive the lowest priority. 

Conversion Attribution Examples

Conversion Attribution in Campaigns Conversion Attribution in Flows Conversion Attribution in Campaigns Vs. Flows

The following examples describe how conversion attribution works in campaigns.

 

Example 1: A customer purchases a product after receiving and acting upon a few campaigns from various channels.

Consider a Push campaign notifying the customer about an upcoming lightning deal, an Email campaign nudging the user to explore deals and offers on the website, and an In-app campaign displaying a banner for ongoing deals and offers. All three campaigns have the conversion goal of a product purchase and have the same attribution window of 36 hours.

John is a user who: 

  • Receives the Push Notification on Day 1 at 4 PM.
  • Clicks on the In-app message on Day 2 at 2 PM.
  • Receives the Email message and opens it on Day 2 at 10 PM.
  • Purchases a pair of sneakers on Day 2 at 11 PM.

LIM_Campaigns_Example1.png

John's deepest interaction is with the In-app campaign because he has clicked the message shown. Thus, going by the deepest interaction, the priority for attribution in case John purchases a product within Day 2 is In-app > Email > Push. Given that the deepest interaction is with the In-app campaign, the conversion is attributed to it.

 

Example 2: A customer subscribes to a newsletter after receiving two campaigns - a Push campaign and an In-app campaign.

Consider a Push campaign notifying the customer to subscribe to receive updates about the latest products and offers from the brand and an In-app campaign nudging the user to explore deals and offers on the website by getting regular updates about offers and promotions. Both campaigns have the conversion goal of a subscription from the customer and have an attribution window of 24 hours.

John is a user who:

  • Views the Push Notification on Day 1 at 10 AM
  • Views the In-app message on Day 1 at 4 PM
  • Subscribes to receive updates from the brand about offers and deals in the app on Day 2 at 8 AM.

LIM_Campaigns_Example2.png

The interaction depth is the same for both the Push and In-app campaigns because he has viewed the messages. We then go with the latest interaction for conversion attribution. John's latest interaction is with the In-app campaign, and the conversion gets attributed to it.

 

Example 3: A customer books a flight ticket on a web page after receiving a Push campaign on their Android mobile device.

Consider a Push campaign for Android customers, notifying them about deals on flights. The conversion goal for the campaign is a flight ticket purchase, and the attribution window is 24 hours.

John is a user who:

  • Views the Push Notification on his Android mobile device on Day 1 at 8 AM.
  • Purchases a flight ticket from one of the web pages that he gets redirected to from the app.

This example highlights cross-platform attribution when a user receives communication from a campaign on a specific platform but completes the conversion goal (within the attribution window) of the campaign on another platform. In this example, the communication is sent using a Push campaign configured for the Android platform, and the user achieves the conversion goal on a web page.

If the purchase event is sent from the Web platform to MoEngage and if there is no Web campaign running with the same conversion goal, this conversion is not attributed to any campaign in the All Interactions Model. In the Last Interaction Model, however, this conversion is attributed to the Push campaign.

 

Example 4: Conversion Attribution when a user falls into the Control Group for two campaigns

Consider two campaigns: a Push campaign and an SMS campaign sending information about flight cancellations. The conversion goal of both campaigns is users viewing the train info is the conversion goal for the campaigns. The Control Group is 5% for both campaigns. For more information, refer to Control Groups. The attribution window for both campaigns is 36 hours.

John is a user who falls in the Control Group for both campaigns. Since he is a part of the control group, he does not receive the campaigns. If John were to view the train cancellation information, he is said to have converted. In such a case, whichever campaign has the latest run gets the attribution. 

 

Revenue Tracking

Revenue Tracking is optional and is associated with the primary conversion goal added to the campaign. For information about enabling revenue tracking for a conversion goal, refer to Revenue Tracking.

If Revenue Tracking is enabled during campaign creation, Revenue Metrics are tracked and displayed on the Campaign Analytics Page. The following revenue metrics are available:

  1. Total Campaign Revenue - the sum of the total order value across conversion events attributed to the campaign.
  2. Average Order Value - Total Campaign Revenue/Number of Conversions
  3. Average Revenue Per User - Total Revenue/ Number of Unique Conversions

Revenue metrics are tracked for all attribution types. Change the attribution type using the filter on the top right of the Campaign Analytics page and view the respective revenue metrics.

Note: In the All Interactions Model, conversions might be attributed to multiple campaigns with the same conversion goal in case their attribution windows overlap. Thus, the revenue is added to all the campaigns with the attributions.

In the Last Interaction Model, only one campaign gets the attribution (the campaign with the deepest and latest interaction), and thus, the revenue gets attributed only to the campaign that gets the attribution.

 

Was this article helpful?
2 out of 2 found this helpful

How can we improve this article?