Can Topics Enable Personalization at Scale?

With Topics, Chrome determines several thematic interests, like “Fitness” or “Travel,” based on a user’s recent browsing history. Topics are calculated and stored entirely on a user’s device without communicating with Google’s servers. They are designed to be transparent and viewable by the user, as well as easily interpreted by potential advertisers via a JavaScript API. This makes Topics a future-proofed solution for simply and clearly providing user’s browsing interests for site owners and advertisers, without exposing detailed browsing history or cookie-based identifiers.

While Topics are aimed primarily at enabling advertisers, it could help with better onsite actions using tools like Google Optimize, Adobe Target, Optimizely, or other onsite testing and personalization platforms.

Brands Need to Look Beyond the Third-Party Cookie

Currently, most optimization or personalization tools cannot be used directly to customize onsite content based on interests. If they can, they are reliant on data management platforms (DMPs) and other brittle solutions that continue to require soon-to-be-deprecated third-party cookies. 

For example, Google Optimize cannot be used directly based on interest or affinity information inferred from third-party cookies. Instead, interest data provided by users who consented to third-party cookie tracking is currently available in Google Analytics for reporting and analysis purposes. 

Since this data is not available for use in audiences shared with Google Optimize, one is unable to target or personalize onsite experiences based on it. Alternatively, while Adobe Target may be able to access this data through API integrations with DMPs, it is also dependent on soon-to-be deprecated third-party cookies, making the data integration moot.

Chrome’s implementation of the Topics API meanwhile presents a ripe opportunity for personalization tools to leverage user interests, agnostic of the third-party ecosystem. In particular, a user’s interest topics will be stored entirely on the browser and accessible via API, meaning that they can be surfaced to a website or other API caller without reference to an opaque third-party cookie. 

This change enables accessibility of users’ Topics directly within the browser, rather than having to be processed on Google’s servers—meaning that a tool like Google Optimize will be able to query and read a user’s Topics in real time when they land on a website, leading to better onsite personalization in a first party context on the browser.

How Topics Can Enable Post-Cookie Targeting

Imagine a user visits our site www.ecommerce.com. As they land on the site, a query to the Topics API returns a relevant interest for that user is “Fitness.” We can surface that topic to an optimization/personalization platform (e.g. Optimize or Target) for targeting, resulting in the optimization platform changing the homepage banner to direct that user to our Fitness product category. 

In this way, Topics and a personalization engine could be used to:

• Update landing page copy and imagery to be more relevant
• Personalize a homepage banner to hero relevant product categories
• Feature content or articles more relevant to a user’s interests 
• Feed personalized recommendations to users

Pending the exact specifications of how Topics are eventually rolled out, even more advanced use cases may be made possible via storing and linking a user’s topics with a first-party identifier. A first party identifier could be cookie based for anonymous users (e.g. a cookie ID like Google’s client ID), or an identifier linked to a known/authenticated user (e.g. a customer ID, a hashed email address etc). In the above example, it may be possible for the user’s “Fitness” topic to be tracked against their first party identifier in an Analytics tool (e.g. stored alongside a customer ID in Google Analytics). 

Once associated with a user’s first-party Universally Unique ID (UUID), this topic could then be integrated with CRM data and used to power relevant marketing automation emails. The topic could also be used to generate more relevant retargeting audiences (e.g. via customer match) or be used to power dynamic creative. Personalization to that level now seems like a hygiene factor with a recent McKinsey report sharing how 71% of customers  expect at least a basic level of personalization. However, a fine line needs to be maintained in how data informs the creative process rather than dominates it. 

Prepare for Cookie Deprecation Now

Topics are still a very new proposal from Google that has been released in order to prompt discussion and feedback, meaning that the actual Topics API has not been thoroughly tested by partners and stakeholders yet. While we can speculate on use cases and possible applications to personalization, all of the above will have to be tested thoroughly as the Topics API is rolled out by Google. 

While we await more detail and further evolution of the Topics API proposal, businesses can partially execute on these use cases by deploying a platform to test and personalize on their website, as well as targeting affinity/interest audiences either via media targeting or profiles available via data management platforms. This would mean a proactive approach to adapt to the eventual deprecation of third-party cookies. Topics may represent a way forward for providing more relevant customer experiences (both onsite and advertising) in a more robust, privacy-centric way.