Use of Audience IDs set to rocket to almost two-thirds of online ad spend by 2020
Cross-device audience ID matching will be used in 58% of total UK online ad spend by 2020, compared with 28% in 2016, according to a new report from Yahoo and Enders Analysis. These ‘audience IDs’ are the online identity profiles created from anonymous identifiers in order to recognise and match the same user across different channels, devices or both.
The report, People, not devices: Audience buying in a Cross-Device World, shows that ad spend based on audience IDs will almost triple in the next few years, reaching €7.9 billion by 2020, compared with €2.7 billion in 2016. The trajectory to growth will see this ad spend reach €4.33 billion in 2017, before hitting €6.22 billion in 2018 - a 226% increase from last year.
When the General Data Protection Regulation (GDPR) comes into effect in 2018, growth in the volume of ad spend which uses audience IDs will slow down, as the advertising industry works to ensure compliance with the new regulations. It will, however, continue to grow, as the industry responds and adapts to the new regulatory requirements.
By 2020, the proportion of display ad spend that uses audience IDs will be slightly higher than for paid search, accounting for an estimated 68% of display spend, compared with 55% of paid search. Most of this growth will come from mobile display and mobile search growth.
The report attributes a large portion of future growth to the recognition from advertisers that using audience IDs will give them a better picture of who is seeing their ads across devices, the ability to better understand and manage frequency exposure and more accurate information about total campaign reach. In addition, cross-device ID based advertising allows for improved attribution, as these IDs can be used to attribute purchases and conversions to show concrete ROI for campaigns.
Nigel Clarkson, managing director, Yahoo UK says “In a world where the number of connected devices per person is increasing at pace, cross-device campaigns should be the norm. Yahoo is already enabling these kinds of campaign, with datasets across search, email and social that allow us to accurately match 89% of users across devices. This level of insight benefits consumers as well as brands and advertisers as it helps get the right message in front of the right audience, making it a win for all.”
Matti Littunen, senior research analyst, Enders Analysis adds “Providing an interface between media, marketing, retail and services, audience IDs are at the heart of the data economy and increasingly important for the online consumer experience. In advertising, they have promise in addressing perhaps the single biggest issue of a changing industry: how to accurately attribute value to ads in a complex online media market? 2017 will be the year media owners, advertisers and agencies come to terms with the full significance of audience IDs.” ()
The use of audience IDs allows marketers to get closer to the notion of having a ‘single view’ of the consumer. Data from loyalty programmes, payment technologies, beacons, ecommerce and in-app trackers can be connected with the IDs used to serve ads, and attribute purchases to prior ad views, further bolstering the importance of audience IDs in the coming years according to the report.
The appeal of cross-device IDs is further increased by the problems of fraud and non-human traffic in online advertising. While not a complete solution, cross-device audience buying at its best can help distinguish between real and fake ad views, and help test claims of ad effectiveness.
Consumers also stand to gain from the growing use of audience IDs. A more sophisticated understanding of the consumer will mean individuals are shown more relevant advertising, that’s less intrusive, as advertisers will be better able to tailor both the creative and delivery methods. This will also mean lower overall ad loads, as the ability to better measure the number of times a consumer sees a particular ad across all of their devices will enable ‘frequency capping’.
Definition of cross-device audience buying:
On a practical level, the shift towards audience buying consists of a number of related trends, referred to in the industry with terms such as “cross-device”, “cross-channel” or “people-based” advertising, or audience buying. While these three have slight differences, they all share the use of online identity profiles or “IDs”, which can be used to identify the same user across channels, devices or both.
We use the term “cross-device audience buying” to refer to all advertising which makes use of IDs advanced enough to identify the same audience segment on multiple devices with a high degree of accuracy.
As was noted by Enders’ Programmatic advertising in the mobile era [2016-018] report, the mobile revolution in online usage has not only made this type of ID more necessary for online advertising, but enabled them to be more robust regardless of device. This makes the growth of cross-device audience buying so important for the industry: as we will outline in this report, it changes what is possible in both the planning of advertising as well as the attribution of value to it.
Types of cross device identifiers:
Cross-device identifiers fall into two main categories – deterministic and probabilistic:
Deterministic identifiers refer to known, ‘people-based’ information that can be linked to a specific profile. These anonymised identifiers are already in use across the major ad platforms, including Facebook Ad Manager, Twitter Ads, Google AdWords, and Yahoo Gemini and BrightRoll.
Probabilistic identifiers refer to speculative information, drawing on behavioural models drawn from different types of data. These data sources are then used in concert to map probable identity across devices. There are still some challenges in probabilistic modelling, as well as the difficulty in encouraging collaboration across the industry. Greater sharing of information is needed between advertisers, media organisations, and middlemen, in order for verification measures to be effective.
