Unlinked audiences

What are unlinked audiences? 

Unlinked solutions rely on publishers and their environment to allow advertisers to target relevant audiences. In this section, we will explore Seller Defined Audiences (SDAs), as well as contextual and attention solutions. Available solutions differ in their approach to how targeting parameters are produced, however all solutions can exist in parallel to linked audiences or an unlinked approach can be used as stand-alone solution. A core benefit is that an unlinked approach will be viable regardless of what happens in the regulatory space in relation to third-party cookie alternatives. Tools include…

Seller Defined Audiences

Seller-Defined Audiences (SDA) is a solution that was created within IAB Tech Lab’s Project Rearc initiative. It is a technical specification that outlines a method for first-party data providers - including publishers and DMPs - to communicate the audience profile of a user to an advertiser, in a privacy secure way and without using identifiers.

SDA is an exciting solution for publishers as it enables them to maintain ownership of, and to monetise, their valuable first-party data.

Here’s how it works:

1. A publisher or data management platform (DMP) creates audience segments based on their first-party data, and each audience segment is given a segment ID
   • Metadata about the segment is provided via a Data Label, which includes information such as the source of the data, data recency, consent mechanisms used, and more
   • IAB Tech Lab offers a Data Transparency Standard Compliance Program that works to third-party verify the data labels, which instils trust from the buyside
2. The segment ID is then sent to the demand side platform (DSP) via OpenRTB
3. DSPs can then choose whether to bid on that impression and audience

Considerations for SDAs

By using SDAs, sellers have more transparency and control over their own data, which can lead to more trust and collaboration between publishers and advertisers. The quality of SDAs depends on the quality of the first-party data that sellers collect and segment, which can vary widely depending on the publisher's data strategy, scale, and capabilities.   

Contextual targeting

This is an age-old technique where advertisers target ads towards people based on the content or characteristics of a web page or environment, rather than user data. This means that audiences are completely anonymous via a pure contextual approach. You also get enhanced contextual environments, which include factors such as the sentiment of a web page or the physical context of the user. Contextual targeting tools include GumGum, Adsquare, Seedtag, Covatic and Integral Ad Science. Find out more with our guide to contextual targeting here. 

Considerations for contextual targeting

As pure contextual targeting doesn’t rely on any personal data, cookies, identifiers or device information, advertisers don’t need to be concerned about compliance with data protection laws - although be aware that enhanced contextual strategies can use some forms of personal data. 

Contextual targeting can also be very effective, with studies suggesting that context is a key factor in how an ad is received and performs. However, frequency capping of ads, creative personalisation and measurement are more limited via a pure contextual targeting approach. 

Attention

Attention is a subject that is generating lots of debate within our industry – from how it should be defined to utilising it alongside other metrics. The premise is axiomatically that ads that are not seen have no impact. Measures such as reach/frequency and viewability are indirect proxies for whether users had seen the advertising.

Attention solutions generally fall into two categories: proxies derived from engagement and proxies derived from eye-tracking. Engagement proxies use browser behaviour to augment viewability such as mouse motion, or the potential time an ad was viewable to create a predicted attention metrics. Eye-tracking proxies use camera tracking to measure ‘eyes on the screen’ in either a natural panel or laboratory setting. A combination of tags measuring browsing environment and eye-tracking data are fed into models to predict eye-tracked user attention.

As both approaches do not rely on any user identifiers, attention solutions can exist in parallel or as standalone of linked approaches.   Attention solutions can be incorporated into campaign workflows through combinations of creative measurement, media measurement and campaign activation.

Considerations for Attention

As Attention is a nascent offering, there isn’t a universal understanding of what constitutes it and how to best measure it. Often attention is deployed in a test and learn framework alongside existing measurement strategies to understand the value. Broadly, the more consumer attention your communication yields, the more likely you’ll achieve shifts across the purchase funnel. Find out more about Attention as a metric.

Summary of unlinked audiences

• Unlinked approaches do not require the exchange of any user identifiers between the various parts of the ad tech ecosystem
• SDAs leverage publisher segmentation and OpenRTB to communicate first-party audiences to the demand side
• Contextual approaches leverage the content being browsed to provide targeting signals to increase relevancy between advertiser and user
• Attention is an emerging field using browser interactions or eye-tracking data to move beyond viewability as a proxy for ad impact