Reaching consumers in a cookieless, privacy-driven world

A decline in addressable inventory is one of the biggest challenges facing brands today. Third party cookie deprecation across browsers, declines in app identifiers on iOS, and regional consumer data privacy legislations are all factors. They've all required brands to explore new approaches when running ad campaigns.

ID solutions such as hashed emails from authenticated users are being used, but they only account for the ‘known’ side of the identity equation, and certainly aren’t the silver bullet to all problems. Today, many publishers authenticate no more than 25% of their users, if at all, leaving a significant amount of non-addressable inventory across the open web.

We forecast that by late 2024, 75% of ad opportunities across the open web will have no identity attached to them. In fact, that number already stands at 30%¹. Under these circumstances it’s not surprising that many brands are turning to alternative targeting solutions that don’t necessarily depend on whether a user is known to them or not. 

A variety of non-addressable methodologies are being explored across the industry, some more established than others. At the browser level we are seeing a new set of features being developed, the most obvious being Google’s Privacy Sandbox initiative for Chrome, which aims to facilitate online advertising without the need for third party cookies. At the publisher level we are seeing the emergence of Seller Defined Audiences, which would allow publishers to match their first-party data with external demand partners, without the need to share any user identifier data. And, of course, there is renewed interest from brands in the power of contextual and advanced contextual solutions to reach users based on their content choices.

But there’s more we can do besides just considering what page URL a user is on or what content they’re engaging with. Panel-based solutions are re-emerging, which aim to project learnings from a sample set of high-quality data onto traffic from unknown audiences. At Yahoo, we’ve developed a suite of Next-Gen Solutions that can help marketers and publishers achieve their goals, whilst respecting consumer privacy choices.

Our Next-Gen Solutions leverage publisher content and other real-time data signals available in each bid request such as weather, time, location and device type. These signals are fed into a machine-learning model, which is trained against a truth set of first-party, consented Yahoo users. The model is then able to accurately infer audience characteristics on a bid request such as age, gender and interests—all without the need for cookies, mobile app IDs, browser storage or creation of user-level profiles. These same machine-learning solutions also enable our DSP to manage ad frequency on a probabilistic basis and bid effectively across ID-less inventory.

And whilst traditional and advanced contextual solutions are certainly enjoying a renaissance, it’s the combination of real-time signals and proprietary data we cross-reference those signals against that go beyond page URLs and content. This data, gathered from our direct consumer relationships, is what makes Next-Gen Audiences a highly accurate and effective solution. Across a number of verticals, brands have already been able to intelligently tap into non-addressable environments thanks to Next-Gen Audiences, whilst maintaining performance at cost-effective rates.

But where do we go from here? Over the coming quarters we want to put more control in the hands of advertisers, to support how they plan, activate and report back on their cookieless activity. For more information on how you or your client can start exploring our available Identity solutions in your market, please reach out to your local Yahoo sales rep or contact us.

Sources:

¹Yahoo Internal Data, 2021