AdTheorent launches Relationship Targeting
The New York City ad tech firm looks to influence in-market buyers by also directing ads at friends and family.
AdTheorent’s visualization of its Relationship Targeting
Many data providers have gone way beyond collecting info about individuals’ behaviors and background, toward mapping the social graphs of that person’s housemates, friends and relatives.
Ad tech firm AdTheorent is putting that social graph to advertising use with today’s launch of its Relationship Targeting.
The mapped social configurations, SVP of Product Management and Business Intelligence Matthew Groner told me, include household, relatives and friends.
Devices that communicate from the same IP address every evening, for instance, might be taken to be in the same household and possibly the same family. Occasional but regular visits to other IPs might indicate relatives or friends, and so on. Relationship maps are assembled via AI that adds, combines and assesses data from customer relationship management systems, mobile location tracking, purchase data and other sources.
The New York City-based firm says it has over 740 million unique device IDs mapped across the 92 million US households, and it is able to determine over 1.3 trillion connections — an average of 14 connections per person.
As an example, AdTheorent pointed to the use case of an unnamed automotive advertiser, who wanted to increase top-of-funnel consideration for consumers looking for a new car.
Positive brand messages
Through Relationship Targeting, ads for the automaker’s new car model were directed at the social circle of an in-market consumer, who was deemed to be in-market because of visits to a car site or similar indicators. Users are identified down to a household level and remain anonymized.
The idea is that ad-exposed housemates, friends or relatives might mention the new model to their friend or family member, who they know is looking. Groner said the social graph ads are often positive brand messages, such as the value of a particular new car model.
In-market individuals who had six or more identified connections — all of whom were served ads — outperformed the desired action (such as clicking on the ad or visiting a showroom) by 33 percent, he said, compared to in-market shoppers with zero to one connection who had been served ads.
Groner added that, while other ad tech firms and data providers are determining social and familial relationships, many of them depend on data from social media, which he says can be deceiving because all “friends” are not always actual friends.
Others, he said, focus on broad-brush IP address targeting, whereas AdTheorent’s relationship mapping is derived from many data sources and mapped via AI, thus yielding a higher-quality understanding of targeted users’ social graphs.
About The Author
The subtle ways advertisers target audiences - and why they shouldn't
Here's why The New York Times broke down its martech and engineering silos
Why media buying and advertising is ripe for artificial intelligence
From algorithms to advertising: 7 steps to introducing AI to marketing