What is Protected Audience API and why is not only about retargeting

The Protected Audience API is a component of the Privacy Sandbox initiative developed by Google. It aims to provide advertisers and marketers with a privacy-centric solution for group-based targeting in digital advertising without relying on third-party cookies. 

In proposed new system, advertisers work with users’ browsers to show specific ads they’d like to display and how much they’d be willing to pay for placements. An algorithm in the browser then decides what ad might appear.

Google’s Protected Audience API will be more than standard retargeting. The algorithms anticipate what kinds of products a user will be interested and have them ready to show before they visit a site. Advertisers can then target for user profiles and protect user privacy rather than tracking specific users. 

How do behavioral targeting and upper-funnel applications relate to Protected Audience API

From the outset, the Protected Audience API, renamed Turtledove and then FLEDGE, was immediately associated with retargeting applications, but later, Google indicated that it’s not the only use case for this API. It’s a powerful tool for building interest groups together with well-crafted reporting tools. Tools such as these will be essential post cookies. By redefining the protected audience API in this way, this solution could be used in a different context. Not only for retargeting, but also for upper-funnel applications as a behavioral targeting technology.

Behavioral targeting is about reaching an audience with common, specific characteristics. This could include types of content consumed online, time spent on articles about specific content or other similarities visible in data. On the other hand, upper-funnel applications, like branding or simple prospecting, are about the goal they aim to achieve: creating a strong brand identity and emotional connection with a wide group of potential and existing customers to foster long-term recognition. 

Protected Audience API in practice

With the Protected Audience API, it’s possible to create interest groups based on data from a single domain by the domain owner or a delegated entity. They’re then available for use on any PA API-compliant inventory across the internet, accessible to the group’s owner.

A typical use case is the creation of an interest group on the advertisers' websites, as they can include detailed information about user activity on the site, such as what products a person clicks through or the value of those products. This data is then used in the ad auction following the Protected Audience API workflow, to place a stong bid and deliver a personalised ad. For example, a person who has recently browsed TVs, might see those specific TVs they browsed, along with other, recommended products from different categories, like a soundbar, a wall mount or even seemingly unrelated products like a fridge—as long as algorithms see patterns.

However, interest groups can also be created on publisher websites. There are multiple valuable data sources about the user accessible from the publisher’s domain, even without the use of third-party cookies. These may include the context of articles the user typically reads, sections they visit or even more nuanced information like time spent on specific content. There are also other data sources that come from the browser, such as user’s topics from Topics API, general geolocation, among others. For example, for a user who has been reading a lot of articles about the best family SUVs, a MarTech vendor might show a video advertisement for luxury SUVs. 

Publisher-based interest groups show great potential

While the value of Protected Audience API in retargeting is undeniable, and multiple companies on the market, such as Google, Criteo, and RTB House, of course, publish lots of material about it, we believe that its upper-funnel applications show just as much potential. 

The main reason is that they don’t require a user to visit the advertiser’s website before displaying a relevant ad to them. All that matters is what kinds of content they’re interested in online and later on they can see ads related to their preferences. All in line with privacy guarantees built within the Privacy Sandbox from Google Chrome.