The demise of third-party cookies offers an opportunity for the digital advertising industry to rebuild on a better foundation. This is particularly true for publishers who are willing to leverage the right approaches and technologies to monetize their audiences and protect their data.
When Google announced its deprecation of third-party cookies in Chrome, advertising technology providers, industry consortia, and Google itself started to work on alternatives. The goal is to build solutions that ensure addressability without compromising privacy compliance.
Finding a valid alternative to the cookie is particularly important for publishers and ad tech platforms operating in the Open Web. Google and other Walled Gardens can count on billions of authenticated users to deliver personalized ads. However, the rest of the industry needs to find alternatives that enable them to address users efficiently to stay competitive.
Publishers can already see what a non-addressable internet looks like. In Safari, where third-party cookies are already blocked, media owners see their CPMs decrease by 50% as compared with Chrome.
Two popular approaches to identity
Today, there are two popular approaches to solving the identification challenge in the post-cookie world. One is based on cohorts and the other uses pseudonymous universal identifiers.
The cohort-based approach
Google has been working on its Privacy Sandbox. This collection of proposals is aimed at preventing individual user information from being shared with the ecosystem. The initiative focuses on local data processing. The goal is to provide technology platforms with APIs to collect aggregated data about user profiles, as well as aggregated campaign performance data. According to Google, the mission of the Privacy Sandbox project is to “Create a thriving web ecosystem that is respectful of users and private by default.”
Grouping users in cohorts may give the illusion of compliance because, arguably, you cannot individually address people. But this is not the real problem. This approach prevents publishers from engaging in a real conversation with people about the value exchange between their data and the services they receive. Moreover, it doesn’t offer transparency and control to consumers. They have no way of knowing in what group they have been added and why. They also lack the ability to remove themselves from a cohort.
The approach based on pseudonymous universal identifiers
The other method available is based on pseudonymous identifiers that are created when a user opts in to share some pieces of information with the publisher or authenticates on a website. This information can be used as a consistent identifier by all the websites that have collected and passed along the advertising value chain. Brands can use the ID to collect information, deliver messages and measure the performance of campaigns.
Pseudonymous identifiers can be created using different types of information and require users’ consent to comply with data protection regulations. When publishers can provide signals such as hashed email addresses or login IDs, these can be used to anchor consistent identifiers across the websites that have collected them.
Most of the time, email addresses and login IDs are not available. With this approach, probabilistic algorithms use passive identification signals, such as IP address and the device’s user agent string that are shared via the HTTP Protocol. This enables the ability to infer the uniqueness of a user across websites. This method can be particularly useful to address and monetize users that are not ready to authenticate yet but are willing to share some level of information with the website.
Why one approach is better than the other
Unlike the Privacy Sandbox, universal identifiers work in all browsers, not just Chrome. Thousands of publishers that are keen to monetize their cookie-less traffic on Safari today and in all browsers tomorrow have already adopted universal identifiers. Chrome’s Privacy Sandbox, on the other hand, is still a set of proposals at their infancy stage. And, so far, they’ve only been tested by Google.
So why would you sit still and wait for Google to develop and test its Privacy Sandbox when universal identifiers are already available and working? Even more so, why should publishers and the rest of the industry rely on an alternative that will further increase its dependency on the tech giant and, most likely, work on Chrome only?
By partnering with the right identity providers, publishers monetize cookie-less traffic in Safari today and prepare for the post-cookie world.
Choosing the right universal identifier
So, if you decide to try the universal identity approach, the first step is to choose which ones to use. As of today, there are over 25 different identifiers that publishers can test in preparation for the cookieless world. No publisher will have the bandwidth and resources to try them all. So, below are some questions and considerations that can help to select the most suitable options for testing.
Privacy and transparency
Does the identifier use privacy-by-design technologies to capture consumers’ data privacy preferences? And does it give consumers the option to opt-out in the future if they decide to revoke data processing access? Make sure that your identity solution provider can guarantee your users’ privacy protection and control over their data.
What about your data? What mechanisms does the identity solution provider have in place to ensure the information that you’re sharing in the bid stream is only accessed by your authorized monetization partners? Data leakage was one of the main concerns with third-party cookies. Ensure that your identity partner can safeguard your and your users’ data.
Footprint and adoption
How many platforms have adopted the identifier? An identifier is useless if ad tech platforms are not using it. If you’re considering several identifiers, verify they have enough footprint to provide some results. Most user ID modules are available on Prebid. (See how many platforms have adopted each of them.)
Cross-domain linking methods
What methods does the identity solution provider use to link IDs across domains? Most of them use deterministic methods and are only able to link authenticated users. No matter how many logged in users and email addresses you have, you will always have unauthenticated users visiting your website. So why miss on the opportunity to monetize that audience if they are willing to be addressed through passive identification signals?
Prepare today to thrive tomorrow
There are only a few months left until cookies are deprecated by all browsers. If you haven’t started testing universal identity solutions yet, start now. You can already see what the cookie-less world looks like in Safari so use this to your advantage. Utilize Apple’s browser as your testing ground and work closely with your monetization partners to understand what solutions bring the best results and why.