The Expert’s Guide to Cross-device Conversion & Attribution

The Expert’s Guide to Cross-device Conversion & Attribution

What is Cross-device Attribution?

Cross-device attribution is the measurement and reporting of conversions as they take place after multiple engagements across different devices. For example, a user may click on an ad on one device, but will finish the desired action on another device. According to our own research, the average consumer will often use 3 different devices before completing a conversion and may switch between screens 27 times per hour. A site that does not use cross-device tracking for measurement is going to have flawed reporting. According to Google Adwords, about 90% of consumers start an activity on one device and finish it on another. Cross-device analytics allow an organization to link a user’s behavior across devices to monitor and optimize the user journey prior to a conversion.

What is Cross-device Attribution?

Cross-device Conversions

Cross-device conversions happen after multiple interactions across devices and sessions, but how does one account for the user journey across different devices and data from walled gardens? Cross-device attribution is the process by which different channels are analyzed and given credit for a conversion that has taken place, as opposed to standard last-touch attribution, which is the default in Google Analytics.

Other Common Types of Attribution

First-touch attribution: The channel that drives the first engagement with a company’s site will receive 100% credit for a conversion, excluding all other channels or points of engagement.

Last-touch attribution: The channel that drives the last engagement with a company’s site will receive 100% credit for a conversion, excluding all other channels or points of engagement.

Position-based attribution: The channels that drive the first and last engagement with a company’s site will each receive 40% credit for a conversion, while all points of engagement via other channels will equally split the remaining 20%.

Linear attribution: The channel that provides an engagement receives equal credit (i.e. four engagements would yield 25% for each channel).

Time decay attribution: The channels that provide engagements closest to the conversion receive more credit (i.e. in a four engagement conversion, the fourth channel of engagement would receive more credit than the third channel of engagement, which would receive more credit than the second channel of engagement, and more credit than channel of the first engagement).

Why is Cross-device Attribution Important?

Using standard analytics and other attribution models may provide an organization with incorrect data that one may need in order to make the best marketing decisions and improve ROI. Cross-device attribution, on the other hand, provides the truest analytics of how a user interacts with a brand, online and offline, regardless of device. The data received can help marketers to better understand the performance of digital marketing strategies and campaigns.

In today’s competitive world, interpreting analytics correctly can mean the difference between profits and losses. With unified data across devices, your organization will be empowered with the data necessary to make marketing decisions that will help you achieve more success with every dollar invested.

How does Cross-device Attribution Work?

There are three main ways to analyze a user’s journey across devices: deterministic matching, probabilistic matching and a hybrid approach of deterministic and probabilistic. 

Deterministic matching involves following a user with a unique identifier. The most common identifier is an email address that is used to log in to websites and apps. For example, Google estimates cross-device conversions by simply following users who are signed into Google services. A user who is signed in to their Google accounts on their laptop, tablet and phone is very easily tracked. While this is one of the most accurate ways of reporting on user behavior, deterministic matching only works when a user is logged in to a given service, meaning that the potential audience involved in deterministic matching is often smaller than other ways of advertising and reporting. 

Probabilistic matching uses anonymous signals to try and identify a user. These signals include:

  • Location (IP addresses)

  • Date

  • Conversion Type

  • Device IDs

  • Landing Page

  • Interests and web history

Probabilistic matching also uses data sets so that the algorithm will be able to utilize machine learning and identify users across devices based on their behavior. Though often less precise than deterministic matching, it often can be applied to a wider audience than deterministic matching can.

What are the Advantages of Cross-device Attribution?

Attribution from any other model than cross-device can be misleading and cause your business to invest in marketing inefficiently. This is where cross-device attribution comes in, to measure exactly what was clicked on, when it was clicked on, and on what device. This enables marketers  to get a complete view of how visitors interact with their site. Other types of attribution may provide you with reporting that falsely provides or denies credit to the proper channels during a user’s often multi-screen journey.

In today’s increasingly mobile-first world, it is vital to make sure that users are engaging with a brand in an optimal way regardless of how many times they switch devices.

How Cross-device Attribution Can Improve Return on Investment

Without cross-device attribution, a marketer would not know that a customer clicked on their mobile ad first before later completing purchase on their desktop. This information is vital for advertisers and companies so that they know where to invest more of their advertising. This type of attribution helps companies to find where their advertising efforts should be focused for maximum impact.

What is Cross-Device Tracking?

Cross-device tracking describes how advertisers, publishers and others analyze a user’s journey across their multiple devices. The goal is to know that a person using their smartphone is the same person that switches to a laptop and/or tablet without sharing personally-identifiable information (PII).

Cross-device Privacy Concerns

The US Federal Trade Commission (FTC) conducted on workshop regarding this issue back in November 2015. In January 2017, the FTC issued a report to follow up the workshop. In this report, the FTC defines what cross-device tracking is and acknowledges the benefits to marketers saying how it is, particularly useful and valuable to advertisers. In addition to highlighting the benefits, the FTC created four recommendations that should be followed by any entity that is using multi-touch marketing. Failure to follow the recommendations will be in violation of FTC Act.

1) Transparency

Transparency, meaning that all entities using cross-device tracking should fully disclose their abilities to consumers. They should provide meaningful information as to whether or not consumers would want to share have this data be collected by companies. Another aspect of transparency mentioned by the FTC was that companies should make truthful claims of the data categories that are collected. For example, data categories could be anything from email addresses and or usernames, these categories are especially important because can include a user's full names. Even hashed emails can be used to re-identify users in some cases. Therefore, it is important to disclose this information to the consumer.

2) Choice

Companies must offer consumers a choice as to how their cross-device activity will be tracked and once chosen it must be respected. This section also states that customers are free to use sort of opt-out tool as they please and that companies may not try and track customers who use these tools. In one case, Turn, an online advertising company continued to target consumers through multi-device tracking despite the consumer’s choice of being opted out.

3) Sensitive Data

The FTC recommends that cross-device tracking should not acquire any sort of sensitive data. This can include health information, finances, children’s information and geolocation information. Unless there is “affirmative express consent” (FTC report) by the consumer all entities must refrain from collecting and sharing this information.  

4) Security

The FTC Act concludes with an emphasis on high security from the data collected from cross device tracking. This is of course required so that there are no incidents of unauthorized/unexpected uses of data. FTC states that companies should only keep necessary

Many believed that monitoring activity across devices would be a real privacy concern. However, the FTC was quick in understanding and addressing the issue to ensure that companies knew how the data should be used and for the general public to understand that at the end of the day they have control over how data is collected and is the type of data that is collected. Google states that consumer privacy a high priority saying that, “only aggregated anonymous data is used” in calculations.

Tapad’s Device Graph™ and Cross-device Attribution

Tapad’s proprietary technology, The Device Graph™, aggregates device IDs from wide-ranging sources such as exchanges, app download data and ID syncs with our many partners. This enables us to deliver a wide array of coverage to our clients and the corresponding ‘network effect’, which provides exponential scale for every new partner contributing to the graph. 

The Device Graph™ sees more than seven billion signals per day, which ensures the graph is never over-leveraged on any one data source or partner. In fact, no partner represents more than five percent of the graph. As opposed to vendors who rely on black boxes or walled gardens, Tapad believes in an open ecosystem and portable data to drive client success.

Tapad helps clients achieve their media execution and analytics goals by licensing access to The Device Graph™ via real-time API and/or batch file delivery. Agencies, brands, ad servers, DSPs and other adtech/martech vendors leverage the Tapad Device GraphTM for use cases that include targeting, acquisition, measurement, multi-touch attribution, churn analysis, customer loyalty, site personalization, customer journey optimization and more.

We also offer a suite of audience products that allow clients to optimize media buying towards conversion, CPA, precision and other key metrics.