To meet the challenge, Oracle Data Cloud leveraged several data assets including CPG transaction data, sourced from 74MM U.S. households, and cookie-based activity/interest data from JavaScript on more than 15MM AddThis tagged websites globally. The datasets were connected via the proprietary Oracle ID Graph™ so relationships between spend and online behavior could be observed and modeled.

With this depth of data and linkage capability, Oracle Data Cloud can model buyers within specific countries. Here’s how it works:

1. Create a seed from CPG transaction data that exhibits desired purchase behavior

2. Limit the seed to those cookies originating from AddThis tags within a target country

3. Apply machine learning to identify cookies similar to those in the seed within the target country

Oracle Data Cloud validated this methodology in countries where transaction data exists finding, on average, Global PBT audiences correctly identified buyers 3x better than non-PBT audiences.

Using the Global PBT modeling methodology previously detailed, Oracle Data Cloud built three audiences to meet the Kraft Heinz Soup campaign objectives. Each seed audience included IDs who made specific purchases and were active in the UK. The model then scaled the audiences to find additional UK cookies exhibiting similar behavior.

1. Canned Soup Buyers – Buyers of canned soup served as the seed audience as Kraft Heinz looked to grow consumption of their Classic Soup varieties among those who regularly purchase canned soups.

2. Natural & Refrigerated Soup Products Buyers – Buyers of refrigerated and natural soup products served as the seed audience for the Kraft Heinz trial objective as they looked to grow share with their new carton soup product.

3. Soup Category Buyers – Buyers of the soup category were added to achieve scale and ensure the message about the Kraft Heinz Soup brands reached anyone with interest in the category.

The Global PBT audiences were compared to non-purchase based audiences served during this same campaign; Global PBT consistently overachieved on key performance metrics.

Nick Barlette, Digital Investment Director for Starcom, was pleased with Oracle Data Cloud’s Global PBT and highlighted the “very strong performance” on video completion rates, viewability, and reach. Video completion rates averaged 75-80 percent for the Global PBT and viewability exceeded the EMEA industry average (60 percent viewability vs. 55 percent – Source: Moat EMEA Industry viewability benchmark).

Based on this initial success, Kraft Heinz, Starcom, and Amobee expanded their global partnership with Oracle Data Cloud, approving 13 Global PBT audiences for Heinz Infant, Heinz Beanz, and Heinz Ketchup programmatic video campaigns. Additionally, Starcom recognizes Oracle Data Cloud as a trusted global advisor, including them in its UK Brand Build day and sessions in France, Germany, and Spain to educate internal teams on the ways Global PBT can help satisfy marketing objectives in any country.