The Predactiv People Model Represents a New Class of AI Foundation Models

Predactiv today detailed how marketers, agencies, publishers, and data owners can benefit from the Predactiv People Model, an AI foundation model built for marketing, introduced last week.

“Predactiv has been known for its data and technology. The Predactiv People Model brings the two together in a new way,” said Dana Hayes Jr., CEO of Predactiv. “It makes our Custom Audiences even stronger and gives marketers new ways to find, understand, and target the people who matter most to their business.”

Trained on data unique to Predactiv
The Predactiv People Model brings together Predactiv’s proprietary open-web behavioral data with other powerful datasets to learn, from a holistic range of signals, the underlying patterns that predict behavior. The Predactiv People Model’s behavioral foundation draws from more than 65 billion unique behaviors every month. 

Read More – AI-Powered Subscriptions Are Transforming One-Time Buyers into Lifelong Customers

What can marketers do with it?

• Unique audiences. Describe an audience in plain language, such as “people likely to buy premium running shoes in the next 90 days,” and the Predactiv People Model builds it from real behaviors, tailored to each request. Since no two clients get the same audience, marketers won’t be over-targeting the same group of consumers.
• Automatic modeling. The model generates a rich embedding for every user, which removes the feature-engineering step that has made predictive modeling slow and expensive. Lookalike, next purchase, and custom models can be built and deployed in minutes instead of weeks.
• First-party data, multiplied. Data owners can onboard their own data, understand who their customers are compared to the general population, and expand seed audiences into modeled, activation-ready audiences.
• Insights in plain English. Because the model’s output can be translated into text, teams can ask questions like “What else do buyers of this brand tend to do?” and get answers grounded in real behavior.

Read More – How AI Engagement Drives Higher Retention in Online Retail

“Lookalike modeling, audience creation, and insights were the first products to run on this model, and the results were telling,” said Michael Gorman, Chief Product Officer at Predactiv. “Clients can now describe an audience or even just upload a brief to get a unique audience that their competitors cannot access. Instead of purchasing static audiences, advertisers can describe the characteristics they are looking for, or upload a seed audience, and have the system produce a custom, high-performing audience in near real-time. That only works if the technology understands people. Now it does.”

Write to us [⁠wasim.a@demandmediaagency.com] to learn more about our exclusive editorial packages and programmes.