Cross-device targeting, the ability to serve targeted advertising to consumers across multiple digital devices, is an emerging technology challenge and will be a hot topic this year and beyond.
Cookies and mobile IDs (UUID, Apple’s IFDA, etc.) allow advertisers to track and serve targeted advertising to uniquely identifiable users but only on a single device or even on a specific app where the ID or cookie is available. Broadly speaking, cookies enable targeting on desktop and laptop browsers and mobile IDs enable targeting on smartphones, tablets, and apps.
Like many of us, I use multiple devices to access the Internet each day. I use my smartphone to check email on the way to the office, my laptop during the day at work, and my tablet to watch videos on YouTube in the evening. Although I am one person, the current device-based targeting technology means that to advertisers and tracking companies I am viewed as three different “users.” A recent TNS study in Australia showed that for those owning at least three Internet devices, share of time was split as follows: PC/laptop (52 percent), smartphone (29 percent), and tablet (19 percent) (Source:TNS Mobile Life 2013).
How Does Cross-Device Targeting Work?
To track a single user across multiple devices, there are two main techniques that can be used:
1. Isolate and link user data from various data collection sources (e.g. connect cookies with a UUID);
2. Use models to accurately infer a link between multiple data points and identify it as a single profile.
The easiest way to link data from various data sources is to have users log in through a “single sign on” (SSO). Google is a prime example here – whether you’re accessing Gmail on your phone or searching for a gift on your laptop, Google knows who you are because of the opt-in login they use across all of their services and they’re able to target advertising and content to you accordingly.
Not all sites are able to get their users to continually log in, though. In fact, very few advertisers will be in a position to have their audience log in via their own SSO (many currently use Facebook).
In the absence of log-in data, tracking users across devices is a complicated process that involves the collection and processing of many different signals. A signal is a data point (such as machine configuration, operating system, or wireless setting) that when combined, can be used to build a unique user profile or “fingerprint” of the device that is being used.
Correlation analysis of behavior across multiple fingerprints can provide an estimation of the “match rate” or likelihood that separate fingerprints (devices) are of the same individual profile. Currently, some services such as Experian’s AdTruth offer tools that help to bridge mobile ID to Web, but full matching across all multiple devices is still elusive.
Why the Hype?
Cross-device targeting promises two core benefits: the reduction of wasted impressions and more effective audience engagement and attribution.
When we rely on cookies or mobile IDs in isolation, each device or mobile app is viewed as an individual user. As frequency caps are set at the user level, it’s possible that as an individual I’ll be exposed the maximum number of impressions on all three devices.
The idea that the number of ads being served to any individual can be capped regardless of where those ads are displayed is a compelling idea for advertisers. Less wastage will lead to two outcomes: the campaign will be more efficient enabling cost savings or the campaign will be able to reach more people therefore driving increased revenue.
The second promise of cross-device targeting focuses on the idea that over time, messages and offers can be delivered to individuals based on their unique profiles. The more data points contained within the profile, the more tailored the message or offer can be and the more accurately it can be delivered.
For advertisers, having the ability to deliver a specific offer to an individual at a given time, regardless of the device that they’re using, is very attractive. It moves campaign planning away from being about mobile versus display and back to a place where the customer is the focus. When the customer is the focus, campaigns will always be more effective.
Finally, cross-device tracking cleans up the attribution challenge of advertising to a single user across multiple devices. Imagine if you’re in the office and you’re exposed to an ad for an airline promotion. You click, you have a look, you start the buying process, but you don’t complete the purchase. You then remember in the evening about the promotion and go on a different device to complete the purchase. In this example, the sale will not be attributed to the ad although it played a key role in driving the conversion. To track this transaction accurately, cross-device capabilities are required.
How Will It Impact Advertisers in Asia?
Asia leads the world in access to mobile broadband. As of 2013, Asia had approximately 895 million mobile broadband subscribers; America had approximately 460 million subscribers and Europe had 422 million. (Source: ITU World Telecommunication /ICT Indicators database.) Mobile access to Internet makes cross-device measurement challenging, as mobile users typically access the Internet via a combination of smartphone, tablet, dongle, and personal hotspot. Even those who access via a smartphone will access via multiple apps, which largely need to be tracked separately. The pattern of Internet access and usage further complicates the tracking and targeting environment.
This raises an important question: When multiple mobile device ownership (phone, tablet, laptops) is on the rise, how will advertisers be able to tap into the value that cross-device targeting tantalizingly promises?
As audience targeting is still in a nascent state in most Asian markets, advertisers should focus on learning to more effectively identify, segment, and target their existing audiences as opposed to focusing on emerging technologies that are still being rolled out. This expertise will prove invaluable when true cross-device targeting becomes ubiquitous in the coming years.