“What’s in a name?” William Shakespeare’s famous line now has a new meaning with respect to Privacy Sandbox, an initiative led by Google to protect user privacy while giving companies the tools and insights they need to better build digital experiences. In studying recent announcements from Google about their new Privacy Sandbox feature called Topics, I’ve noticed the new feature will have implications for how brands claim their space online—in particular, it may be time to consolidate under a single domain name.
What’s topical about Chrome’s Topics?
The Chrome browser currently has a strong market foothold with above 50% in market share, despite not really having a solution following Safari’s (Apple’s) crusade to kill third party cookies. While it hasn’t gone for marking the cookiecalypse yet, like Safari and other browsers have, Chrome has attempted to address the needs of marketers with ideas to alter some marketing capabilities to work in a privacy-safe way.
Google has sought solutions which would enable some form of safe profiling and data exchange between martech players as ad revenues continue to be mission-critical for the health of many businesses. These solutions are going to be built into the Chrome browser, packaged as the Privacy Sandbox.
The most recent announcement introduces Topics, which are an updated version of FLoC (federated learning of cohorts). The initial idea behind FLoC was to create a mechanism which would classify users based on their behavior into cohorts which guarantee privacy (through entropy). By design, cohorts would be more generic and would remove 1:1 targeting, but at the same time would restore interest-based targeting. The main issue of the initial solution was in mathematics, as algorithms were translating domain names into numbers without a clear understanding of “topicality” of the site.
As FLoC did not win hearts of the industry, Google went to the drawing board and came back with Topics. The idea of cohorts still persists, though the mechanism of translating domain names into “topics” for further targeting was updated along with some privacy assumptions. Even though the proposal is not yet fully developed, there is a consistent approach of using domain names in order to classify users into cohorts. Google envisions some form of a dictionary and set of rules that determine which domain name translates to which topic. Current documentation points out that usage of sub domains is encouraged to support mapping into topics.
How does the third-party cookie crumble crush my current domain name strategy?
The main issue with third-party cookies is that they enable “foreign” actors to collect information about individuals as they travel between different sites. Though as everything in a binary word of computers, the definition of a foreigner is very black and white. All cookies set by a different domain are considered third parties. Computers do not care much about the structure of your organization, brands, subsidiaries and ownership.
Multibrand businesses which operate across multiple domains face challenges in building user profiles without third-party cookies. As data management platforms (DMPs) and many marketing solutions struggle to exist without third-party cookies, it is becoming more difficult to create a single customer view across brands one may own. As first-party data strategies are picking up steam, there are some critical decisions to be made. To operate in a first-party cookie context and be able to exchange data between their own brands, brands need to operate under the same domain name.
The most apparent manifestation of this situation are media outlets which own multiple mastheads. As publishers try to build value propositions around their audiences, every piece of information counts. Without a DMP or third-party cookie, it would be impossible to achieve scale across different sites they own today.
So, how can I make a name for myself?
As it is possible to register your own top-level domains or TLDs (though they are expensive and it takes time), and we observe ongoing pressure on first-party data collection (meaning you need one universal domain across your whole business), it's time to consider your new universal domain name!
Let's assume you run a business called “Example” together with two brands, “Big” and “Small.” It’s likely you have example.com, big.com and small.com as domain names. With the lack of third-party cookies, it is hard to exchange information about prospects between the sites. With help of a customer data platform (CDP) or a good data team, you may join first-party data between the sites to research the level of cannibalisation or overlap.
To simplify your life (and data), you may want to consider big.example.com and small.example.com as primary addresses. This will enable all sorts of integration and will load your first-party data strategy with rocket fuel. If you are big enough, you can go for your own top-level domain to create something like big.from.example and small.from.example. Coming back to Topics, if your brands operate with multiple categories, more subdomains enables better profiling, like automotive.big.from.example or sport.big.example.com.
How do I get started now?
Well, FLoC did not survive long enough to become a thing and Topics are still quite nascent themselves. Though everyone is pretty committed to getting rid of third-party cookies, and some businesses already operate in a world where over 80% of traffic comes from browsers that no longer support them by default. Google has postponed the moment of putting the final nail into the cookie’s coffin, so the timelines seem rather floaty.
Today we operate with the assumption that hour 0 will come around next year or the year following. All things considered, there is not much time to prepare for such big decisions. Now, it’s time to plan.
Make our digital heart beat faster
Get our newsletter with inspiration on the latest trends, projects and much more.