Three Groups That Should Plan for a Privacy-Forward Internet Now


For several years, privacy has been a key topic of discussion in the world of adtech. This increased dramatically in the past 12 months, in line with Google’s intention to deprecate third party cookies from its Chrome browser later this year. Despite Google’s recent delay of this timeline to 2025, groups throughout the industry are still rapidly proposing and evaluating a myriad of techniques meant to increase internet users’ privacy while also preserving the precision of digital advertising. 

Although development and testing of these solutions is far from over, one notion is already clear: from types of identifiers to data clean rooms, the industry has not and will not choose a unified path forward. 

Among this industry-wide debate, Google’s Privacy Sandbox initiative has proposed the most prominent solutions, the three most discussed being:

  • Protected Audiences (PAAPI): serves remarketing and custom audiences without the use of third-party cookies. The most apt substitute for 3PCs.

  • Topics API: designed to preserve privacy while allowing a browser to share information with third parties about a user's interests based on recent browsing history.

  • Attribution Reporting API (ARA): enables measurement in a privacy-preserving way, without third-party cookies.

However, conversation surrounding this initiative’s proposed APIs focuses largely on what they can do, not what it will take within the advertising ecosystem to make them work.  

While the industry laments the heavy lift at a macro level, there is not enough discussion about what it will take to be ready at the micro level. The changes proposed by the Privacy Sandbox will require significant effort for marketers and advertisers, publishers, DSPs, data providers, and analytics platforms. This disruption will be too much for some adtech players to adopt on their own, creating the need for additional service layers. In order for the industry, and the web itself, to have a strong future, we need more players across the industry to get involved in testing and adoption. 

Before we look ahead to the changes of the Privacy Sandbox, it’s important to remember how a recent disruption, Apple’s AppTracking Transparency (ATT) and SKAdNetwork, impacted the industry. Prior to these releases most DSPs and Social Media companies were able to easily attribute conversions to in-app inventory. After ATT, attributing conversions to in-app inventory became far more difficult.  Mobile Measurement Partners (MMPs) stepped up to provide services to help fill the attribution created by ATT. These MMPs didn’t have clear solutions from day one. It has taken years of development to work within these new systems. Will we see the same type of adoption process with Privacy Sandbox? Will new types of companies and offerings arise to serve both advertisers and publishers? 

Ensuring advertising is both private and precise requires more sophisticated technologies than the industry leverages today. Most of these APIs don’t function in a silo and require supporting technologies or intermediary steps. For example, to get the most out of PAAPI, you’ll likely need to set up a Trusted Execution Environment (TEE) and a Google-Provided Key Value Store, which requires an additional TEE. These two pieces alone could take a team of staff engineers multiple weeks to stand-up and additional time to test and validate. If you don’t have a highly experienced team, they may need to learn new coding languages and concepts causing further barriers to entry. 

And that’s just half of the puzzle. Reporting and attribution will be an additional number of weeks, if not months. While the set-up can be prohibitive, it does push the industry toward a new level of privacy. Let’s look at a few uses cases and where different players in the ecosystem may run into specific problems:

  • Publishers

    • While the publisher PAAPI auction set-up might be a smaller lift, the attribution component can be complex and noisy. It is not a plug and play solution. 

    • Publishers may require a partner with advanced knowledge of numerous attribution techniques in order to get an understanding of their site’s performance without 3PCs

  • DSPs

    • DSPs will require innovative tactics to maintain targeting, activation, and measurement. While the Privacy Sandbox offers solutions, if a DSP isn’t prepared to jump in with both feet, the technical development could take months or years. 

    • Operating on PAAPI requires shifting targeting and bidding to the browser, requiring a rebuild of all related systems.

  • Agencies

    • Brands often rely on agencies to deliver performance, which often comes through 3PC targeting (retargeting & prospecting) that is being limited by industry changes.There’s now only one main path to activation and reporting on Chrome. Agencies may need to be prepared to make tradeoffs in order to service their clients at scale.

Given these disruptions, all players in the industry should be prepared to answer difficult questions of whether to build, buy, or partner. There are numerous areas that contain technical nuances that some publishers, DSPs, and agencies may not know how to handle. As an industry, we need to help these different areas of our ecosystem prepare and easily adapt to these challenges. Here are a few action items teams will need to consider: 

  • Infrastructure Support 

    • Think about the effort required to set up TEEs and other aspects of propping up the most advanced solutions, such as a Key Value Store, within the Privacy Sandbox

    • Even an experienced team will need to deal with specific challenges, such as standing up Infrastructure as Code (IaC), sending data to TEE services, and creating jobs to send data to the TEE services and asynchronously retrieve the results

  • Identification

    • Testing both deterministic and probabilistic identifiers as browsers make updates to what is permissible

    • Updating core components of cross-device graphs

      • Be prepared to hear the term “ID Bridging” more than you can stomach

  • Attribution and Measurement

    • Moving from individuals to cohorts via ARA and other solutions

    • Enhancements to conversion or event modeling 

    • Certain Privacy Sandbox services allow a report to be run only once. It’s important to work out design decisions and implementation issues early, and test with small datasets, since the raw data becomes useless after the first run. This is a major risk for those who are inexperienced. 

  • Audience Creation

    • Nearly every audience a brand is able to activate  today is going to change besides 1PD

    • Creating and activating based on interest groups

This list will likely grow as solutions development continues, requiring extensive support from groups that will need to dedicate time to addressing these areas, both in the short and long term. 

Shaping the future to provide better privacy while maintaining precision and performance in advertising is a difficult task for the industry. Whether it’s considering new first-party data strategies, or investing in the Privacy Sandbox, every company needs a path forward. Those who aren’t asking those questions today are already behind. 

Don’t be caught sitting on the sidelines. Companies that are willing to test new technologies and service layers will be ahead of the curve in the world without 3PCs. While the ecosystem will evolve greatly over the next several quarters, early adopters with an agile mindset will be the winners. The time to test is now.