Skip to main content

The rise of cross-device attribution

You may start reading this on your laptop at work, then you get busy with a million other things and have to stop. So you pick it up later from your tablet while you’re at home. Or maybe from your phone while you’re on the treadmill at the gym. In today’s world, we all have multiple devices that we use at different times of the day. Marketers call this cross-device. Users call it the norm. The challenge however is more than just creating content and advertising that can live at any size and on any device, it’s connecting the dots between the devices to determine accurate attribution.

Currently, marketers have a larger array of tools at their disposal, but even with this ever-growing solutions arsenal, it can be very challenging to understand which marketing activities are driving growth. With the spread of connected devices per user, from the attribution perspective, it has become more challenging to get the big picture. The difficult reality is that this device connectivity evolution has made us start questioning the validity of metrics that were once trustable. Digital marketing attribution models allow us to see which activities are driving positive change, but can we pinpoint which type of attribution is more relevant: cross-channel or cross-device attribution?

There’s no question that both are important. Cross-device is part of cross-channel attribution for a very important reason: understanding that multiple interactions can come from the same person automatically enhances your data. It provides a broader and clearer view of your marketing data. With these two models so tied to each other, how do you make the best use of them?

Differences between cross-channel and cross-device attribution

Let’s begin by understanding the variances between the two subjects. Cross-channel attribution explains which media, emails, and search keywords are the most effective, assuming that users could have been across all of them. Cross-device attribution provides information related to the experiences of each consumer across different devices. Cookies are essential for these two to work; however, cross-device attribution goes beyond by revealing that a single user can be represented across multiple cookies. Implicitly, cross-device attribution is a subset of cross-channel attribution. The next step, after understanding what campaigns are having a positive influence, is to understand the number of times a user interacted with each campaign. A key part to accomplish this is to be able to track users’ behavior across multiple devices.

How cross-device attribution has evolved

A few years ago, the desktop was the only device where ads could be delivered. What this meant was that you could always safely assume that if your campaign had a reach of 100 users, you were more certainly sure that all 100 unique users were touched by your campaign. But as we all know, that is no longer the case today, where users may be exposed to the same message multiple times on multiple devices and across many different marketing channels. A potential scenario from this environment is that a user may visit or purchase on a website on one device, but they could have easily been exposed to the campaign on a different one. In general terms, this level of difficulty did not exist until recent years.

What this challenge brings is a disconnect between what those metrics are that were once valid and the accuracy of what they mean in today’s scenario. A consumer with two devices may look like two unique users in our data.

What we need to do is redirect our thinking towards thinking about people and devices when analyzing our data. In the near future, the term “unique user” may be an obsolete term. It’s very important that in order to achieve cross-channel attribution while considering cross-device data, we need to understand that “users” and “devices” no longer have the same meaning.

So, what’s next for attribution?

So what do marketers need to do to get a more holistic view of how consumers reach their buying decision? A basic premise is that they need to fully understand cross-channel and cross-device attribution and their synergies that can be tied all together with a method called “data stitching”, which can be a tedious process. One of the roadblocks that marketers come across is the fact that many mobile devices do not support cookies, compared to traditional reporting solutions that use cookies to track consumer behavior. Nowadays, companies are still working on getting a grip on this new methodology.

A potential resolution on the horizon is the data-management platform (DMP). Once companies get a tight grip on DMPs to successfully gather and compile first-party deterministic cross-device data, the more chances they will have to carry out the marketing initiatives with a more holistic picture. What this can potentially allow is to move customers through the sales funnel with a sequential, more targeted, messaging approach. The use of unique user-generated IDs has started to become more popular and it’s helping marketers identify their consumers and their activity across devices.

A huge competitive advantage can be achieved by being able to message a single user across multiple devices, which in turn, can provide a better experience for the user. Being able to determine that a user’s path to conversion included three digital channels is a step in the right direction, but the giant leap in attaining a complete attribution picture is in being able to determine that a user’s path to conversion included three digital channels across three devices. Attribution is more efficient when data at the most granular level is complete and accurate.

Cross-channel and cross-device reporting are critical in order to have an efficient data-driven marketing plan. At a higher level, cross-channel attribution can provide data that can be acted upon quickly, thus it is paramount that marketers are ready to drill down into cross-device to have a more complete attribution picture. The premise is that we have to look at these two attribution aspects and their intertwining as an opportunity rather than a complexity that cannot be leveraged.