New York, NY – January 18, 2022 – AdTheorent Holding Company, Inc. (Nasdaq: ADTH) (“AdTheorent” or “the Company”), a programmatic digital advertising leader using advanced machine learning technology and privacy-forward solutions to deliver real-world value for advertisers and marketers, today announced the launch of its AdTheorent Predictive Audience Builder, a transformational suite of platform tools designed to enable users to create and activate predictive models which score audience quality.
AdTheorent Predictive Audience Builder leverages customizable and primary-sourced seed datasets to mimic the audience profile of an advertiser’s desired target. In a major departure from industry-standard audience segments, that seed data set is not used for direct targeting. Instead AdTheorent’s machine-learning reads signals from those data sets to build a predictive model which scores programmatic inventory on its likelihood to reach an individual who meets the desired profile. This privacy-forward predictive scoring delivers superior audience quality and KPI performance without the use of personally identifiable information, cookies, or IDs of any kind.
AdTheorent is proud to include automotive brand Southeast Toyota Distributors, LLC and 22Squared as a launch client.
AdTheorent Predictive Audience Builder’s use of primary-sourced, highly-customized audience profile parameters is boundlessly flexible and customizable to each advertiser’s marketing strategy.
· Examples include completely customizable vertical-based audiences such as: auto intender in-market for a specific make or model; frequent fast-food diner with high probability of switching to a new chain, or frequent family meal or online orderer; big box and family shopper with high household income; or high spender on luxury travel in market for or researching a trip.
How It Works:
· Primary-Sourced Data: AdTheorent Predictive Audience Builder leverages primary-sourced datasets (provided by either AdTheorent or an agency or brand) to identify audience quality statistics relevant to the specific brand campaign. Examples of data types include:
o Consumer Data: Thousands of consumer data attributes such as demographics, purchasing habits, lifestyles, interests, and attitudes.
o Location Data: Precise location data sourced directly from in-app SDKs and server-to-server integrations with publishers and mobile application developers.
o Verticalized Data: Vertical-specific data across automotive, B2B, CPG, dining, finance, retail, travel and more.
· Machine Learning Expansion: AdTheorent identifies commonalities in the data using machine learning and identifies other important attributes to grow the addressable ML-informed audience in real time.
· ML-Based Audience Optimization: As per AdTheorent’s standard Predictive Targeting processes, AdTheorent Predictive Audience models define the data parameters within which AdTheorent ads are served, with the primary goal being optimizing ad delivery towards data attributes and combinations which cause KPI conversion lift.
· Campaign Performance: Using AdTheorent’s Predictive Targeting, the campaign is optimized toward the advertiser-specified KPI to drive performance. AdTheorent delivers an ad to an impression opportunity only when AdTheorent’s predictive models indicate a sufficiently high probability that a given ad opportunity will do each of the following: (1) be served within the customized Predictive Audience and (2) lead to completion of an advertiser-specified campaign action. Each AdTheorent Predictive Audience model evaluates millions of impressions per second to drive performance, considering 1000+ data variables in its models. Models self-optimize throughout each campaign, allowing AdTheorent to drive industry-leading performance for advertisers.
For more information about AdTheorent’s new solution, click here.
Melanie Berger, AdTheorent