Balancing Your Tactics for Post-Cookie Marketing Relevance

Solutions for continued success without third-party cookies and device IDs

As the industry prepares for a future without third-party cookies and device IDs, an influx of solutions have surfaced to preserve granular targeting and attribution while protecting consumers’ privacy. Simultaneously, investments in this space are on the rise, with U.S marketers spending $7.8 billion on audience data activation solutions last year.

The industry is trying to make sense of the changes in order to implement successful strategies, while balancing numerous unknowns such as market perception and solution adoption. And, of course, also deciding where to invest inhouse vs. rely on partners. To industry experts, it’s evident that consumer-centric solutions built on trust, transparency and consent will be the most successful and sustainable. Moreover, ongoing analysis, evaluation and discovery will help determine which solutions will be around for the long-haul, and which will have short-term relevancy.

To cut through the noise, here’s some clarity on the approaches most likely to help buyers and sellers continue to achieve their goals in a future without third-party cookies and device identifiers.

Publisher first-party data

Publisher first-party data clearly has a place in the long-term future of advertising. The elimination of third-party cookies has given many publishers a strong reason to increase engagement with their users, and many have employed strategies to incentivize users to share their data in exchange for some value, be it access to content or promotions to encourage loyalty. Investment in strengthening direct consumer relationships and transforming resulting insights into meaningful audience data for media buyers will be a differentiator for publishers who do this successfully.

Many publishers who are all-in on a first-party data-oriented strategy also need to consider ways of working with buyers without their proprietary data inadvertently spilling over into the ad ecosystem at large. Ultimately, publishers with effective first-party data strategies will maintain positive user experiences on their properties and respect users’ privacy nominations while successfully monetizing their inventory.

Publisher IDs—which could be domain-specific first-party cookies, or consistent across sites based on authenticated data—are valuable for the purposes of frequency capping across a publisher’s properties and facilitating publisher-provided audiences, often sold through deals. Buyers benefit from being able to leverage the strength of the publishers’ consumer knowledge to reach audiences with relevant advertising.

Industry IDs

While scaled publishers could attract sufficient advertising demand, not every publisher is in a position to rely on publisher ID strategies alone. For publishers who want increased demand access, industry IDs like LiveRamp’s Ramp ID or the Open Source Unified ID 2.0 offer connectivity to buyer data.

Additionally, for buyers with first-party data, forming independent relationships with every publisher is not sustainable. In scenarios where both advertisers and publishers have direct consumer relationships, industry IDs offer compelling possibilities for cross-site targeting, frequency management and attribution in a privacy safe, transparent way.

Industry ID solutions fall into two main categories: deterministic and probabilistic. Deterministic solutions rely on an authentication event. For example, when a consumer provides their email in exchange for some value, and consents for their data to be used for advertising. Such data can be resolved to industry IDs and used in programmatic auctions. Deterministic IDs are cross-device as they represent a user rather than a device and are more durable than third-party cookies. Deterministic ID solutions have the key benefit of transparency to consumers, being anchored on data that is explicitly provided by a consumer to a brand they recognize. However, due to the opt-in nature of deterministic ID solutions, availability is naturally limited.

Probabilistic ID solutions utilize modelling and data science to make inferences that a combination of signals represents a unique device or individual. While identifiers built upon probabilistic methodologies have the potential to facilitate greater reach and cross-site connectivity between publishers and advertisers, they are device-specific, less accurate and less transparent than deterministic solutions. Combined, these impose a level of doubt around the long-term sustainability of probabilistic ID solutions in light of privacy standards and regulations.

Industry ID solutions are rapidly evolving, and we expect those that clearly articulate to consumers the value exchange of content and data will find a permanent place in the ecosystem to provide user-based connectivity across the open web.

Contextual solutions

One of the key limitations of ID-based solutions is the ability to scale. Given some projections that only 20% of inventory will be identified deterministically, contextual capabilities enable scale that we will not see from identity-based solutions.

Of the plethora of available solutions, contextual is the most established and well known. While many solutions are still in their infancy, publishers and advertisers have been and can continue to leverage contextual capabilities offered through their partners today. More recently, many publishers have also begun to leverage their first-party data to provide unique contextual offerings, and the industry as a whole is collaborating on many fronts to align on standardized content and audience taxonomies.

Contextual has advanced from simplistic keyword match targeting to multi-signal analysis driven by machine learning to drive advertising outcomes while avoiding the use of identifiers. Contextual solutions can be employed in parallel with other approaches wherein contextual signals, such as content category, keywords and sentiments, are supplemented with other user-based data to extrapolate insights and infer interest or purchase intent.

While contextual approaches provide high reach, and are widely regarded as privacy-preserving, they do not solve for frequency capping or measurement use cases. This is why contextual approaches are often used in conjunction with other solutions, such as modelled frequency or even publisher first-party ID offerings.

Browser APIs

Browser APIs are among the newest of emerging solutions which could play a role in cross-site targeting and measurement use cases. These proposals—discussed in World Wide Web Consortium (W3C) meetings and published on github—aim to enable classical marketing strategies in new ways while maintaining the privacy and anonymity of the consumer. They achieve this by placing the browser at the center of targeting and attribution workflows.

For example, the Federated Learning of Cohorts (FLoC) proposal works by grouping users into cohorts based on their browsing history, without that browsing history ever leaving the device. Other proposals cover retargeting use cases by creating interest groups stored within the user’s browser (e.g., FLEDGE), or Apple’s Private Click Measurement (PCM) proposal which facilitates post-click attribution across websites, without arbitrary cross-site tracking. Although these proposals are interesting, many are still very much in their infancy. While some are already being tested (e.g., FLoC), many questions remain about their practicality, utility and even their privacy-preserving nature.

Moving forward, it will be imperative to work together as an industry to both protect consumer privacy and leverage appropriate data for media buyers and sellers to succeed. To help publishers and marketers make decisions based on their unique priorities, it’s necessary to invest in a comprehensive approach to support first-party and consumer-consented IDs. These solutions—detailed in Xandr’s recent guide “Shaping The Future of Identity”—help consumers see the value exchange from advertising, making it a better experience for all parties.

Ewa Maciukiewicz serves as VP of product management at Xandr, working across all audience and identity products. Previously, Ewa also served as senior director of product management at Xandr, where she drove product vision and strategy along with the roadmap for audience, identity and privacy products across Xandr’s TV and digital platforms.