Cross-Device Reporting May Be The Most Important Thing You’ll Do With Your Google Analytics Data In 2018

by Hollis Bowman

As some of you may have hung around long enough to recall, way back in 2012 Google began the process of upgrading to Universal Analytics. The promise of Universal Analytics was pretty ambitious – there were exciting changes to the type of data that could be integrated in GA with uploads, the introduction of custom dimensions and metrics, and product integrations with the rest of the Google stack. However, Universal also marked an important (though, at the time, pretty subtle) shift from session-based analytics to user-based analytics. Why is this important? Because for the last 6 years as an industry we have been evolving toward users as the central organizing principle of digital measurement. With user-level data, marketers and digital analysts can uncover insights in cross-session activity, user retention, and conversion rates based on actual people and not on frequency of exposure.

I sincerely believe marketers and digital analysts are finally really making good on user-based analysis across their entire digital experience with a brand. But (and this is a big but) doing so requires a fundamental revolution in how we think about digital analysis. This is not a simple process.  As an industry we need to be ready to move out of using our marketing tooling in isolation. This is simply not reflective of how customers interact with brands, and in user-based analytics there will probably never be one tool to rule them all.

In this article we’ll review how to use GA to make significant strides toward cross-device use. However, to really cross the threshold away from sessions and toward users, marketers need to develop, deploy, and maintain a repository of customer data across tools. User-based analytics, and the reality of brands spanning devices and delivery platforms (website, social, app, in-store), can only be solved by joining this data in a single place, which is where Customer Data Platforms (CDP) come in.

Connecting Devices To Users – A Primer

Many digital marketing tools provide ways to make connections between devices for users on that platform specifically. Typically this is done through a very explicit opt-in by signing in to your account, and is the backbone of most current versions of cross-device reporting. Facebook, GA, Adwords, DCM, and Attribution 360 all have cross-device offerings. There are others to list, but the sea of martech is both vast and foreboding so I won’t even try to list everything. The problem is that the identifiers used to tie together users across devices for each platform aren’t interoperable with other data sets. So I know my users on social, in GA, and in CRM and/or email, but I might not have a particularly reliable way to blend those tables into a comprehensive view that is platform-agnostic.

Below are two methods for deploying user-oriented cross-device tracking. While both have not insignificant development requirements, one is effectively an evolution of existing tracking while the other is likely a significant change to the architecture of all your organization’s marketing data and therefore a bit more complex. Not surprisingly, the most complex is also the best way to do this as it will provide comprehensive user records tying back to identifiers from your numerous digital assets. That type of data, once available, is a powerful tool to manage and execute against current and new privacy regulations and enforcement, and provides you with rich data sets from which to create targeted omni-channel campaigns.

Methods For Cross-Device Reporting

Neither of these solutions is a quick-fix. Both will need significant development time and executive buy-in to be launched successfully. I’m going to break down what each method will produce in terms of value to help you secure those resources and executive sponsorship.

Method 1 : Google Analytics User ID And Cross Device Reporting

Google has provided a high-level overview of the process you need to follow here. Since I get to see a whole lot of data from many organizations, I’m going to review the outcomes for people who choose to enable User-ID so you can decide if this approach is a good fit for your business.

Is it worth it?

Maybe. As much as I believe User-ID views are extremely useful, I personally don’t find the cross-device report suite to be on the top of my value-added list, as these reports don’t give you a lot of flexibility or actionable data.  There are things you can do to make this data more useful, which we’ll get into in detail below.

Requirements:

  • Somewhere for users to log in to an account
  • Updates to your privacy policy to cover this type of data collection
  • Enable User ID
  • Create User ID view
  • Development resources

How to get more out of GA User-ID for cross device reporting

There are two things I recommend that you do to go over and above the default functionality of Google Analytics user-ID tracking for cross-device reporting.

  • Send a User-ID that integrates with other customer data: The default User-ID reporting in GA locks you into a Google-specific view. This limits your ability to follow your users through other databases such as CRM or ecommerce data that contain a wealth of data that, because it is sensitive personal information, is not able to be integrated to Google Analytics. You can make your GA user-ID data an important piece of context in your internal customer data sets with user-ID by making sure the value that is sent in GA for this dimension is either identical to, or translatable with, similar identifiers in your other data sources. This allows you to blend datasets together in a larger database.
  • Send a User-ID as a Hit-scoped Custom Dimension: You can also get more out of Google’s User-ID reporting if you send the value for the default dimension as a custom dimension as well. This significantly increases your flexibility for reporting, allowing you to use this value in custom reports, segments, and API pulls. This dimension should be hit-scoped so that you can apply more granular analysis about page and event interactions on the site.

This allows you to answer questions such as:

  • Do users associated to multiple mobile devices have a higher AOV?
  • What happens if we target direct mail to users who have interacted with specific product content in the last month?
  • What happens if we target email sends to users who have purchased on mobile devices?

Bonus point: GDPR compliance

Perhaps one of the more important pieces here is comprehensive, cross-device and cross-platform user data will make GDPR compliance scalable and comprehensive, as long as you can leverage that data with other internal records so that it is not locked into Google-specific customer identifiers.

General limitations and considerations of Cross-Device Reporting in GA

There’s an excellent support file about this, but here’s the TL;DR on the limits of cross-device reporting in GA that will be particularly meaningful for your ability to use the data:

  • They are only available in a User-ID enabled view
  • User data won’t line up with other reporting views, as you only see users for whom there is an associated ID
    • Because a lot of people create and log in to retail sites in order to make a purchase, and many users are still more inclined to purchase on a desktop than on a mobile device, this can bias your data toward desktop compared to the rest of your reporting, and also toward the most engaged users overall. I recommend that User ID views not be used to compare against your usual reporting view, with the one exception of analysis that seeks to compare behavioral trends between highly-engaged users and the rest of the population.
  • Cross Device Reports are limited to a 90-day lookback from the present day
  • Revenue is associated with the device on which the transaction completed, with the exception of the ‘Revenue from Originating Device’ metric found in the Acquisition Device report

Method 2: Deploy And Integrate User Data On A Customer Data Platform (CDP)

On the face of it, a CDP is similar to a Data Management Platform (DMP) or a Customer Relationship Management (CRM) database. The main difference is that these systems are built and governed by marketing departments. They are a persistent record of your customers, presented in a way that can be used with your other marketing tools. There are a few major players in this space right now, and I’m going to keep this applicable to any and all of them. That said, we have done work for many years with Tealium and their AudienceStream product is definitely on my list of leaders in solutions for this space.  

Is it worth it?

Probably. Deploying a CDP is a significant investment of time and effort, and will require a comprehensive plan for how to deploy, govern, and maintain data. But this effort will provide a significant benefit to the custody, control, and ability to leverage customer data for your marketing department. Unlike the GA-specific solution, a CDP helps make data planning and maintenance less onerous and manual. For example, they make managing records of tracking opt-in or deletion requests more clearly documented and simple to maintain (a huge bonus for GDPR compliance). Once established, these systems are also designed to be used by marketers, reducing your ongoing requirements for IT support.

Requirements:

  • A data plan – what is tracked and what are the standards to which the data must comply?
  • The help of your legal team to release an online privacy policy written to explicitly cover this type of tracking, and a process for customers to opt-in / opt-out of tracking on your site
  • Customer identifier for all platforms – this is the identifier that is platform specific
  • Primary user identifier – this is the identifier that unifies the platform-specific  ids
  • A CDP vendor (check out AudienceStream)
  • Data integration work to share data from other services with the CDP, and to share the CDP’s identifier with other services where appropriate.

Outcomes:

  • Create cross-device behavioral analysis to better target and engage audiences
  • Unify the brand experience for users across platforms and devices to create a seamless experience through the whole customer journey
  • Reduce reliance on IT for data management and compliance
  • Increased governance in campaign management tasks
  • Much more comprehensive records of users to support GDPR compliance, so that if consent to use this data is withdrawn, the customer’s data can more easily be purged from various systems

Wrapping Up

This post started with some of the promises and challenges of cross-device and cross-media measurement. User-based analytics presents a metaphor for measurement that reflects user behavior on the ground.  People don’t restrict their interaction with your brand to your website, or to a single device. Marketers and digital analysts need to evolve to match this behavior to help unify the customer experience across the myriad of ways they interact with your brand. Unified, cross device data will help you position the right message to the right audience on the device and channel that speaks to them best. Taking this seriously will involve a great deal of planning and execution, but will put you in a position of strategic competitive advantage for leveraging marketing data and also provide a scalable and comprehensive approach to GDPR compliance.

Hollis Bowman

Senior Analyst and Google Analytics 360 Practice Lead

Hollis Bowman is Napkyn Analytics' Practice Lead for Google Analytics 360. As a senior member of our Analyst Team, Hollis' specialty is working with our clients to turn questions into data and data into answers.