One of the biggest challenges for digital marketers today is being able to connect the dots as consumers increasingly engage with their brand across a variety of devices in different settings.

These days, it’s pretty common for people to browse items on a website using a mobile phone while commuting to work, look up competitor pricing on a desktop while at work, and then come home and use a tablet to make a purchase.

Using traditional analytics tracking, it’s difficult to make the connections to single customers and the relationships between their disparate gadgets, in order to build the full profile of the user experience. Cross-device tracking makes this possible with three primary options, using either deterministic or probabilistic techniques to connect activity across disparate devices.

Deterministic methods include user authentication and walled gardens, and use system-generated customer IDs. These methods are more accurate than the probabilistic device fingerprinting technique.

1) User Authentication

User authentication is a deterministic tactic that employs the use of specific identifier such as a customer ID, login or other user-specific data to create a link between behavior on different devices.

Say you’re browsing on Amazon.com. When you log in to your Amazon account with your mobile phone or tablet, Amazon.com keeps track of your device and can identify where you are in the buyer journey – even across multiple devices. All of your behavior is trackable using the account ID you’ve logged into.

This is the concept behind cross-device tracking with Google Analytics (Universal Analytics) and Adobe Visitor Identification. It’s an effective tactic, but it’s not scalable because it is limited to registered users or past customers. You can’t track new visitors unless they create an account and you assign them an ID.

The 1st-party source of information means this method is extremely accurate and it’s highly recommended to use this even if you have a low login rate on your site or low return visits. You can always supplement it with one of the other two methods.

2) Walled garden method

The “walled gardens” we’re talking about here are 3rd-party networks with excellent cross-device history on a wide swath of the population such as social networks and telecommunications companies who can provide you with the cross-device graph you are looking for.

Large networks and social sites like Google, AOL and Facebook can identify users across devices using their own user IDs. Some of these networks offer tracking of your customers across both your web properties and theirs, and across devices.

The services they provide are usually in the form of aggregated reports and analyses. None of the networks provide services to read and utilize their User ID for you to use in your analytics platform for cross-device tracking. Sometimes cross-device tracking is offered as part of other marketing services as with Google’s AdWords and Adometry or Precision Market Insights by Verizon Wireless.

Whether you partner with the networks or purchase the information from them, this deterministic option can yield quite accurate cross-device tracking. Although it is a more scalable option than User Authentication, it has limitations. Companies like Google and Facebook cover a larger percentage of the digital population but they are still confined to the size of their user base and will likely limit access to the type of data available for analysis.

3) Device Fingerprinting

Device Fingerprinting is a probabilistic tactic for cross-device tracking, and one that has the most promise for scalability.

Device Fingerprinting employs selected attributes from device settings and browser options which can then be combined with IP addresses, WiFi info, and users’ web browsing behavior to build fairly unique user identifiers for cross-device tracking.

Device Fingerprinting techniques were developed a decade ago to replace website tracking cookies that were not allowed in the earlier versions of mobile browsers. However, these techniques have found new uses in connection with data management platforms (DMPs). DMPs collect data from websites and other data sources in huge quantities. They can then mine this data, based on your organization’s specifications and derive answers to specific questions you have about users including their cross-device behavior. Tapad is a leader in this technology and partners with several DMPs and marketing measurement services. A 
few DMPs (Krux and Lotame) have built their own algorithms.

Again, this method has the potential to be less accurate than the other two options, but it is very scalable. If you’re already considering a DMP for programmatic media buying or other marketing applications, this would be a good option to supplement your 1st-party cross-device tracking, especially for new, unregistered users.

Many organizations will employ multiple methods to strike a balance between accuracy and scalability. No one method is perfect in and of itself, but they get us closer to the end goal of being able to map a customer’s journey from interest in a product or service to final purchase across a variety of devices.

Case in point: Improving Customer Experience with Cross Device Behavior Measurement

Objectives:

• Collect the data required to understand the customer and make improvements against those business objectives
• Gain a holistic view of the customer experience across digital assets and devices

Leveraging dedicated mobile application tracking, we recently worked with luxury online retailer 1stdibs to design and deploy a Google Analytics mobile measurement solution for the iOS app, which included tracking users across not just touchpoints, but the multiple devices they use for a more holistic view of the customer experience. This meant a whole new level of understanding their customers. For example, 1stdibs didn’t know if people would actually buy on their new app or just use it to browse, returning later to purchase on the website. By tying together cross-device sessions with the User ID feature, they can answer these kinds of questions and see how people are moving between the app and the website and what those experiences look like. These translate directly into customer value and help react to and plan for customer behavior. Read the full case study.

Results:

• Informed by analytics, paid media campaigns had a 47% lift in transactions and a 10% gain in overall return on ad spend (ROAS)
• Data-driven approach to email marketing strategies led to a 34% increase in email click-through rate
• With a deeper understanding of purchase path, an aggressive A/B testing strategy & redesign of the Product to Checkout experience led to a 24% increase on conversion quarter over quarter