Google adds Gemini-curated ad bundles to Display & Video 360

AI-curated inventory bundles

Google recently announced that its Display & Video 360 Marketplace now features personalized media bundles curated by Gemini models. These bundles are specifically designed to help advertisers quickly secure space for live sports and major tentpole events. As reported by PPC Land, the update aims to streamline the process of finding high-impact inventory across different formats.

The new feature appears in the Marketplace's 'Featured' tab, which highlights premium content and top-tier publishers. These curated lists are categorized by various factors, including region, audience affinity, and specific formats like CTV, video, and audio. By grouping these assets, Google allows advertisers to see pre-selected deals that align with specific marketing goals.

While the 'Featured' tab focuses on these curated selections, the 'Discover' tab remains a place for broader exploration. The Discover tab lists all available publishers, central partner deals, and private deals in one unified list. This two-pronged approach gives buyers the choice between AI-guided selections and a comprehensive manual search of the entire inventory.

How bulk assignment works

The primary technical change is the introduction of a 'Bulk assign' button within the Marketplace interface. When a user selects a curated bundle, they can choose to assign it to a new line item or an existing one. This allows for the one-click activation of multiple inventory packages, which is intended to be the fastest setup for complex campaigns.

By default, the system is designed to select every deal within a featured list when a user clicks the assignment button. If a buyer wants a more specific selection, they must manually uncheck the packages they do not want to include. This workflow is particularly useful for CTV performance strategies where managing numerous individual deal IDs can be time-consuming.

There is a specific technical constraint on how many items can be included in a single campaign. Each line item can contain a maximum of 50 packages, which serves as a hard ceiling for bulk assignments. If a buyer tries to add a bundle that pushes the total over this limit, they must deselect some of the existing packages to proceed.

AI-powered search and ranking

The Marketplace also features an AI-powered unified search engine that helps buyers find inventory using natural language or specific keywords. The search draws on a wide range of metadata, including publisher descriptions, content categories, and specific genres like action or comedy. It also integrates Google audience segments, such as specific affinity categories like cooking enthusiasts.

A significant detail in the search logic is how the system handles multi-word queries. For example, a search for 'sports fans' will return publishers with that phrase in their description, but it also surfaces packages with high coverage on related Google audience segments. This means a package might appear in results even if the term 'sports' is never explicitly mentioned in the package description.

The ranking of these search results is determined by a combination of relevance and scale. Google weighs how closely the inventory matches the query against the total available impression opportunities. This means that smaller publishers with highly relevant but limited inventory may be ranked lower than larger publishers with more total volume. Google does not disclose the specific weightings used for these two factors.

Impact on ad buying

For advertising agencies and in-house trading desks, these updates represent a shift toward curation over manual negotiation. Instead of building complex spreadsheets to track various publishers and deal IDs, buyers can rely on lists assembled by Google's AI. This is especially relevant for high-pressure environments like live broadcasts where timing is critical and inventory needs to be secured rapidly.

However, this convenience comes with a trade-off regarding editorial control. Because Google's AI decides which bundles and publishers are shown first, the platform takes on a more prominent role as an editor of the marketplace. Advertisers are essentially trusting Google's models to determine what constitutes 'premium' or 'timely' content for their specific campaigns.

This update is part of a broader trend in programmatic buying where inventory is packaged into single tradable units before a bid request reaches the buyer. By hosting this packaging inside its own buying platform, Google is centralizing the curation process. This move simplifies the user experience while simultaneously consolidating the flow of information between publishers and advertisers.