Google Chrome Privacy Sandbox: Future of Data-Driven Advertising
The ad industry is abuzz with the recent announcement that Google is seeking an industry alternative to the “third-party cookie” in 2 years time. From AdAge:
Google says it’s launching an initiative called “Privacy Sandbox,” a collaboration with other industry players to find alternatives to cookies while limiting the fallout to publishers and the rest of the ecosystem. Cookies are the digital tokens that websites leave on people’s browsers to log information about their whereabouts online.
Google says it just started working on new ways to think about serving targeted ads while maintaining privacy. One method could be to keep consumer data in large enough pools of fellow web users so that individuals maintain anonymity, but also share characteristics that would be useful to advertisers.
This move affects only browser-based advertising and not mobile in-app advertising that uses mobile advertising IDs (MAID) or identification for advertisers (IDFA), which are persistent though resettable IDs set by the mobile operating system. The majority of time spent is now on mobile in-app so a large portion of digital advertising is not going to be affected. In addition, with the rise of connected TV (CTV) and over the top video (OTT), there are new device identifiers being used that are not cookie-based and not affected. That said, browser-based advertising is still a material portion of digital advertising and it’s critical the industry solves this problem.
Thunder has been thinking and planning for this future for some time. Here is how Thunder is leading the industry in solving for data-driven advertising and privacy at the same time.
Differential Privacy
With its Truth in Measurement industry initiative, Thunder convened in 2019 dozens of the largest publishers and advertisers to discuss new approaches to protecting user privacy while providing advertisers’ reliable cross-platform measurement. This industry group endorsed an approach by Thunder known as differential privacy, which is fundamentally what Google is only now in 2020 proposing.
Differential privacy is a system for publicly sharing information about a dataset by describing the patterns of groups within the dataset while withholding information about individuals in the dataset. Grouping users in pools is the key to maintaining anonymity. Here’s additional resources on learning the basics of differential privacy:
- WTF is Differential Privacy (Digiday)
- Introduction to Differential Privacy (Thunder)
Thunder is now in a working group with the industry leaders to finalize such a standard that Google and others can follow to allow data-driven advertising to work with cross-device/platform accuracy and user-level privacy. We expect to release such a standard by the end of 2020 to enable Google and the industry to adopt in 2021 ahead of the 2-year Google goal.
Authenticated, Consented Users
Separately, Thunder is collaborating with leading industry players such as MediaMath and LiveRamp on several different methods that use 1st-party cookies, which are different than the 3rd-party cookies being sunsetted by Google in 2-years.
1st-party cookies are cookies set by the website you’re visiting. If you visit CNN.com for example, CNN can set a CNN.com cookie on you that is considered 1st-party while you’re on CNN.com. If a traditional ad tracker such as Doubleclick drops a cookie from Doubleclick.com while you visit CNN, it would be considered a 3rd-party cookie and in 2 years, it may no longer work. 1st-party cookies are essential to keeping users “logged-in” and remembered by the websites you visit.
Going into the future, Thunder’s COO Ka Mo Lau predicts more publishers will require you to register and log-in in order to read their content. You’ll have to opt-in (or opt-out) of that log-in data for being used for targeted advertising. Once you’re logged in, an anonymized person-based identifier can be used and shared with ad tech companies such as Thunder to provide targeted messages to consented users.
Much of the ComScore top publishers are now testing a form of this solution with the different providers and by the end of 2020, a majority of potential Internet reach should be covered by some form of authenticated consented user solution where the publisher enables targeted advertising for its logged-in users.
Going forward
The industry has at least 2 years to accelerate some of its existing efforts. Since the beginning, Thunder has been an advocate for a more privacy-centric, people-based approach. We are looking forward to working with the industry and want to hear from you if you’d like to join us.