• by Laurie Sullivan , Staff Writer, 7 hours ago

AI is changing the requirements for identity, as agents begin making thousands of campaign decisions per second. This means advertisers will need identity resolution and audience signals to work in real-time across every major buying environment.

Samba TV last week introduced "Project Gravity," an independent platform that takes offline consumer lists such as email lists and matches them to digital profiles so they can be targeted with online ads.

"Agentic activation means brands can move from a handful of target customer personas to potentially thousands of them and drive highly personalized messaging that would have been cost-prohibitive without AI," Ashwin Navin, co-founder and CEO of Samba TV, told MediaPost. "Mid-flight optimizations of media mix, vendor allocation, messaging and creative optimization will become faster."

At the core, AI is only as smart as the identity and data that powers it, he said. Good advertising has always relied on human creativity, which will become even more important in the era of AI slop, he said. 

Samba pushes those matched consumer profiles into ad networks to launch a live advertising campaign across more than 40 major ad platforms including Amazon, Google DV360, Magnite, Meta, TikTok and The Trade Desk.

Rather than upload a customer email list separately to 40 platforms, a brand can upload it once to Project Gravity, and the platform standardizes the data and coordinates it across all.

Advertisers must instantly unify consumer data across devices and platforms to fuel automated bidding systems. Failing to synchronize data instantly across environments renders automated bidding ineffective, as it relies on outdated, unusable information.

As AI agents take on more of media buying, a new bottleneck has emerged because even the smartest media-buying system is limited by the speed, quality and neutrality of the data it can act on.

"When data lives inside a walled garden, an agent will optimize for the platform that owns the data, not the brand paying for the media," he said. "The brands that will win in the agentic era will take control of their audience strategy and run it across the platforms. Neutrality is no longer a nice-to-have. It’s critical for brands to take control and not be single-sourced to any one platform."

From the first line of code, the platform was engineered to solve the real-time data needs of the agentic era, according to company documents.

The platform matches advertiser data against Samba’s Identity Graph to enable targeting based on privacy rules, suppression, frequency-capping and audience expansion. The company said it does not own or sell the ad space to reduce any conflicts of interest.

Navin says data needs to support platforms in real-time with accuracy for identity resolution. “We couldn’t find a solution, so we built it,” he said.

Acxiom, a data broker, uses Project Gravity to target audiences across different networks without being tied to specific ad inventory.

Most legacy data matching tools rely on yes or no rules, so if the technology cannot match a consumer to an exact email address, it drops them.

Project Gravity uses AI vector embedding -- a way to turn complex, real-world data such as words, images, or consumer behaviors into a list of numbers, so a computer can understand its meaning.

Rather than looking at exact matches, it allows AI to understand context, concepts, and relationships. Embedding places data into a multi-dimensional map, which makes targeting more exact.

Every piece of data gets an x, y, z to coordinate and elements that share a similar meaning or context are placed close together, which allows for more exact targeting.