- User Guide
- Getting Started
- Setting Up Attribution
Airbridge Attribution Model
User journeys are often complex, stretching over a long period of time. To build successful strategies that can navigate such intricacy, you need to analyze which of your marketing efforts are most effective in driving desired business outcomes. This process is known as attribution.
The Airbridge attribution model is a framework for distributing the credit of conversion among the various touchpoints in a user journey. You can access the attribution data in Airbridge reports to measure ad performance and optimize conversion paths.
At base, Airbridge uses the last-touch attribution model, which assigns full credit to the touchpoint closest in time to the conversion. The touchpoint that most likely led to a desired business outcome is called the winning touchpoint.
However, there are cases where the last touchpoint does not take credit because the Airbridge attribution model prioritizes certain touchpoints over others.
The Airbridge attribution model is applied to the Actuals Report, Trend Report, Reinstall Report, Retention Report, Funnel Report, Revenue Report, and Active Users Report. Other reports use different models that are more suitable for their reporting purposes.
Airbridge identifies the winning touchpoint under the framework of the Airbridge attribution model. Note that each player in the mobile marketing ecosystem has its own attribution rules, sometimes resulting in attribution discrepancies. The attribution rules also vary depending on where the touchpoint occurred.
Web attribution rules
Airbridge’s web attribution rules differ from the mobile app attribution rules in the following aspects.
You can’t set a lookback window for web events.
The attribution window for web events is set to 3 days (72 hours) by default. If you wish to make changes, refer to this article.
Touchpoints collected via tracking links are prioritized over touchpoints collected via Airbridge Web SDK’s UTM parameter parsing.
Users exposed to advertisements respond in various ways. They might decide to install an app, sign up, or even make a purchase.
Airbridge defines some of these desired business outcomes as Target Events and uses them as trigger points for attribution. This means that when the Airbridge SDK detects a Target Event, it notifies the Airbridge server, setting the attribution process in motion.
Some conversions that are not Target Events are referred to as Subsequent Events. Below are more details on the two types of events.
Target Events include App Installs, Deeplink Opens, and Deeplink Pageviews.
App Install is the act of downloading a software application onto a mobile device. The Airbridge SDK detects an App Install when the app is launched for the first time after the download. Reinstalls are also categorized as App Installs.
Deeplink Open is the act of clicking on a deep link to be redirected to a specific in-app location.
Deeplink Pageview is the act of clicking on a deep link in an app to be redirected to another location within the same app.
Remember that Target Events are the trigger points for attribution. The winning touchpoint of a Target Event also takes credit for Subsequent Events that take place within the attribution window, which will be explained later in this article.
Requirements for tracking Deeplink Pageviews
To track Deeplink Pageviews, you need to configure the Airbridge SDK so that you can use tracking links within your app. For detailed instructions, refer to the articles below.
Before you configure the SDK, note that the attribution of Deeplink Opens may be affected if there are too many cases where Deeplink Pageviews take place immediately after Deeplink Opens.
Subsequent Events are desired business outcomes that are not Target Events and that take place within a defined timeframe after a Target Event. Conversions that are outside of the timeframe, or the attribution window, are not counted as Subsequent Events. App opens, purchases, and subscriptions are some common examples of Subsequent Events.
Remember that the winning touchpoint for a Target Event also takes credit for Subsequent Events. For instance, if a touchpoint from Channel A has been determined to take credit for an App Install, the touchpoint takes credit for all Subsequent Events within the attribution window as well.
Having notified the Airbridge server, the Airbridge SDK proceeds to collect data about the Target Event and reconstructs the user journey.
The user journey consists of touchpoints and conversions, ideally associated with the same device identifier. When a Target Event occurs, Airbridge identifies touchpoints that occurred before the Target Event and conversions that occurred after the Target Event.
Below are more details on the user journey, touchpoints, and conversions. For more details on the data collected by Airbridge, refer to this article.
A user journey consists of a series of user interactions with advertisements and the resulting business outcomes. The ad interactions are referred to as touchpoints, and the business outcomes as conversions.
A touchpoint is any interaction between a user and an advertisement. Impressions and clicks are the two essential examples of touchpoints.
Each player in the mobile marketing ecosystem has its own definition of touchpoints. For example, one ad channel might only count an impression when an ad plays for at least 10 seconds, whereas another might consider 3 seconds sufficient.
Some touchpoints are uniquely defined and supported by ad channels. Examples include the "Engaged View" supported by Meta ads and Google Ads, as well as the "Engaged View-through" supported by TikTok For Business.
A conversion is a desired business outcome resulting from advertisements. It generally takes place within your app or website. App installs, purchases, and subscriptions are some common examples of conversions.
Airbridge categorizes conversions into Target Events and Subsequent Events. Target Events include App Install, Deeplink Open, and Deeplink Pageview. Subsequent Events are conversions that are not Target Events and that take place within a defined timeframe after a Target Event.
Not all touchpoints are given equal significance in the attribution process. Above all, fraudulent touchpoints are always excluded.
In addition, if there are multiple touchpoints in a user journey, only those of the highest priority are considered. This means that those of lower priority can never be identified as the winning touchpoint.
In the same context, not all attribution methods are given equal significance. For more details on the attribution methods used by Airbridge, refer to this article.
Below are the Airbridge attribution model’s touchpoint prioritization rules.
From highest to lowest:
Tracking links
UTM parameter parsing
Organic referrer attribution
Airbridge Web SDK
The priority is determined by the user's interaction that generated the touchpoint. Attribution is primarily based on the highest priority touchpoints, starting from priority 1.
Click
Impression
For example, if both a user's ad click and impression occur within the lookback window, Airbridge's attribution model only considers the click for attribution.
Some touchpoints are uniquely defined and supported by ad channels. Examples include the "Engaged View" supported by Meta ads and Google Ads, as well as the "Engaged View-through" supported by TikTok For Business.
In general, the priority of these touchpoints is higher than impressions and equal to or lower than clicks. For instance, an Engaged View from Meta ads has a higher priority than a regular impression. Therefore, if both an Engaged View and a regular impression occur within the lookback window, Airbridge's attribution model only considers the Engaged View for attribution.
From highest to lowest:
Deterministic attribution
Probabilistic attribution
For more details about Airbridge's attribution methods, refer to this article.
For accurate attribution, it is crucial to ensure that the data collected along the user journey are relevant to the Target Event. Time is an important factor here.
The impact of an advertising campaign tends to diminish over time. For instance, if a user installs your app after seeing two advertisements — one from the previous year and the other from the day before, the latter likely had a greater impact. This shows the importance of setting a specific timeframe in which touchpoints can be associated with conversions.
You need to determine the time before and after the Target Event for data collection. The period preceding the Target Event is referred to as the lookback window, and the succeeding as the attribution window. Touchpoints and conversions outside of these timeframes are not considered for attribution. Below are more details on the lookback and attribution windows.
The lookback window is a user-resettable timeframe preceding a Target Event in which touchpoints are considered relevant. Touchpoints outside of this period can’t take credit for conversions.
Let’s say the lookback window is set to 3 days (72 hours). If a Target Event took place at 2:00 PM on Jan 4, Airbridge will only consider touchpoints from between 2:00 PM on Jan 1 and 2:00 PM on Jan 4.
You can differentiate the lookback window by attribution method, touchpoint type, and ad channel. For detailed setup instructions, refer to this article.
The attribution window is a user-resettable timeframe succeeding a Target Event in which conversions are tracked and recorded. Conversions outside of this period aren’t attributed.
Let’s say the attribution window is set to 3 days (72 hours). If a Target Event took place at 2:00 PM on Jan 4, Airbridge will only measure conversions from between 2:00 PM on Jan 4 and 2:00 PM on Jan 7.
You can differentiate the attribution window by Target Event and ad channel. For detailed setup instructions, refer to this article.
Having collected data within a set timeframe along the user journey, Airbridge determines the winning touchpoint according to the rules below. Remember that Airbridge uses the last-touch attribution model.
When a Target Event takes place, the Airbridge SDK collects touchpoint and conversion data within the lookback and attribution windows, respectively.
Among touchpoints of the highest priority, the one closest in time to the Target Event takes credit for both the Target Event and the Subsequent Events. Fraudulent touchpoints are always excluded.
Deeplink Opens and Deeplink Pageviews are prioritized over App Installs. If a Deeplink Open takes place within the attribution window for an App Install, the window is reset to start from the point of the Deeplink Open.
Was this page helpful?