• by Laurie Sullivan , Staff Writer @lauriesullivan, February 20, 2023

Fantix, a company that allows even small businesses to compare first-party data sets with others, has launched a partnership with Epsilon.

The company calls its business model a “federated knowledge engine” and describes it as technology that enables marketers and business analysts link and compare first-party data sets with others for a more holistic view of the consumer journey.

Through the partnership, Epsilon will enable small to medium-sized companies to match their first-party data to Epsilon foundational data assets to gain analytics and insights in a privacy-safe environment, utilizing comparisons as insights resources.

The Fantix application uses data abstraction to fuel authentic, privacy-safe insights.

Abstraction is a proprietary method used by Fantix to aggregate and anonymize data before it leaves a data owner’s computer or server.

Abstractions are one-way aggregations that cannot be reversed.

By only collecting abstractions and preventing access of user-level data, the company said it supports competition by allowing others make data comparisons in a privacy safe way. 

Antonio Tomarchio, executive chair and co-founder of Fantix, founded in December 2021, and former founder of Cuebiq, says the technology sits on the premises in the cloud on the company’s own servers. Companies license the technology.

Fantix does not have access to the data.

“Abstraction is better to use than encryption, because it cannot be reversed engineered,” Tomarchio said.

Advancements in machine learning and artificial intelligence, along with the depreciation of browser cookies continues to make first-party data a critical component in a brand’s advertising strategy.

New rules make it easier for larger companies like Amazon to compete, but what about smaller companies.

Tomarchio believes this is the future of analytics and machine learning, where developers train models in a series of databases without sharing data to learn more about customers.

“We are doing this as a first step in the direction,” he said.

The service works off a model where companies can run one report or sign up for a longer subscription. When an organization begins to use the service, the person uses the Fantix application to create a data abstraction of their own CRM on premises, which is sent to Fantix.

Similar to a data clean room, but much different, according to Tomarchio, users can connect with each other to compare abstractions, but the data never leaves the premise of the data owner.

“In a clean room, you push data into the clean room and have a data partner to do encryption,” he said. “It’s usually one to one or one to a few, but you’re still pushing your data outside of your premises. We do not require the movement of any data at all.”

When two users connect, Fantix calculates the areas of overlap between the abstractions, generating aggregated statistics and insights. Users of the app are only charged when they gain insights and enrich their data, and spend a fraction of what comparable market-research analyses typically cost.

Companies participating in the federation have the ability to determine what and when they will contribute. They can create competitive filers and more.