blog

The Drawbridge Approach to Attribution

Determining the impact of marketing efforts on customer purchases and conversions is no easy task. Successful attribution has become especially challenging in recent years. According to a recent survey, attribution topped the list of digital advertising issues in 2017.

Think about it: not only do customers use three devices per day on average, they can also be exposed to upwards of 5,000 ads per day! With all of those touch points across so many offline and online channels, how can marketers best determine which efforts drove consumer purchases?

Thinking About Attribution Models

Attribution modeling is the go-to solution for marketers looking to better measure marketing impact on customer buying behavior. There are a wide variety of attribution options for marketers. Below are some of the most popular attribution models.

  • First-Touch: The first-touch attribution model delivers all the credit for leads and sales to the first source that touched the customer. So, if the first touchpoint for a conversion is an ad impression on a mobile app, then all of the conversion is credited to the mobile app. 
  • Last-Touch: This attribution model delivers all credit for leads and sales to the last source that touched the customer. If a consumer views impressions across multiple devices and converts, the conversion is attributed to the device and/or channel that delivers the very last impression.
  • Multi-Touch: The multi-touch attribution model distributes credit for leads and sales to all sources that touched the customer. So when a consumer views or clicks ads across a variety of devices, this model distributes the final conversion across all of these touch points that occur within the attribution window.  

Multi-touch attribution can give you the full story of all of the different touchpoints and devices that lead to your customer’s conversion. This model is helpful when you want to understand the consumer’s full purchase journey across all available devices before they make a purchase. 

How Drawbridge Approaches Attribution

Drawbridge has a robust process in measuring conversions and applying attribution modeling.  

Conversion Measurement

Before we discuss Drawbridge attribution methodologies, let’s start with how Drawbridge measures conversions. Drawbridge receives conversion data in two ways: Drawbridge receives conversion data in two ways: 

  1. Web conversions: “Conversion pings” fired from advertiser mobile or desktop webpages
  2. Mobile app installs: Drawbridge receives app installs in the form of postbacks from our attribution partners, including Kochava, Tune, Adjust, branch.io, and Appsflyer. We also receive a flag which indicates whether Drawbridge drove the install or not. Using this information, we stop serving ads to consumers who have already converted from other sources.

Drawbridge measures these conversions using the following process:

  1. We consume conversion pings in the form of unique IDs.
  2. Using our Connected Consumer Graph®, we review impression and click logs within the conversion look-back window to search for extended unique IDs with matching app IDs. (Note that we use the graph version of the conversion time.)
  3. We use app IDs or conversion IDs, which is defined by the advertiser, as unique keys for claiming conversion events.

In the next section, we’ll dive into details regarding our cross-device conversion models.


Drawbridge Attribution Models

Drawbridge offers two attribution models: last-touch and multi-touch.

Last-Touch

For this model, Drawbridge attributes the entire conversion to the true last interaction. This interaction can happen on either the same device or cross-device in the form of a Click or View. Drawbridge always prioritizes a Click above a View, because we assume a Click is more valuable in driving conversions than a View. If a consumer converts more than once, we use two logics: Claim-All or Non-Claim All. 

With Claim-All, we claim all conversions.


With Non-Claim-All, we only claim one conversion.


Time-Decay Multi-Touch

While there are many methods of distributing weights for multi-touch attribution, Drawbridge uses a time-decay model. This model assumes that interactions closer to the conversion are more valuable than those that occurred earlier.


Using Drawbridge for Attribution 

Drawbridge can provide cross-device attribution insights or we can deliver “raw data.” 

Multi-Touch Attribution Report - Cross-Device Insights Suite

This report, part of our Cross-Device Insights suite, is designed to show marketers the number of conversions based on both last-touch and multi-touch attribution. These insights can be leveraged to better update and plan your media campaign channel budgets.


For example, if you take a look at our sample report above, you can see that the multi-touch attribution model provides more conversions to mobile devices, even though they weren’t the last device to touch these converters before they made their purchase. 

This adjustment impacts the CPAs of both the desktop and smartphone channels! In this case, this advertiser may want to allocate more spend to smartphone and less to desktop.

Log-level reporting based on Drawbridge ID (BYO Attribution Model Offering)

The new Drawbridge ID, our proprietary identifier, helps marketers and analysts gain a better understanding of their audience’s journey to conversion, helping to accurately attribute conversion credit.


Based on the Drawbridge ID connections, we can deliver a log-level file that connects ad impressions, clicks, and conversions to the Drawbridge ID, making it easier to see which desktop, smartphone, or tablet impressions are connected with conversions.


You can then overlay your own attribution model on top of this file to determine which of these cross-device touch points should be attributed to conversions. 

Connected Consumer Graph (BYO Attribution Model Offering)

By leveraging the Connected Consumer Graph, you can build your attribution model on top of a rich, fully linked data source. The primary graph format, our cluster file, links our Drawbridge ID to any associated device IDs and cookies. This option is ideal for DSPs looking to improve their cross-device attribution offerings.

Blog

Product Pulse: Fall 2017 Release Notes

Read More >

Social

We get a lot of questions about the best way to measure success and attribute #crossdevice conversions – so we've t… https://t.co/mybO8qX9st

Social

THRILLED to be a finalist in the 10th annual Media Excellence Awards for our #crossdevice product-level attribution solution! Congrats to all of the finalists!

Social

Congratulations to our friends at @MGOMD / @Specsavers on #winning Most Effective Cross-screen Campaign from… https://t.co/uYMHew3z5h