Ad targeting failures—why brands are missing their intended audiences

When brands run digital ads aimed at parents, they may actually be reaching an audience where 67% don’t even have children. And if they’re specifically targeting “moms,” more than half of those consumers could be men. Mismatched ad targeting and faulty data were among the findings of a new survey from ad tech platform Adlook that identified glaring gaps in the programmatic ad landscape, where personalized advertising appears to fall far short of its promise. The recent survey highlighted a few egregious examples where data providers in programmatic ad markets missed basic targeting demographics. Adlook surveyed about 1,300 people online, asking the consumers if the data used to reach them matched their characteristics. In many cases, the data was off. The survey tapped into data from about seven different data providers, commonly known within the ad industry, although Adlook did not disclose the names of the companies. They were “the top data management platforms, which are providing data at scale,” said Mateusz Jędrocha, Adlook’s chief product officer. Data-Driven Marketing Playbook Join us May 22 in NYC to explore how data powers marketing Learn more here Also read: I took a data journey—here are the disturbing things I found Adlook did not name any data providers because the purpose of the experiment was to show there were better ways to conduct digital advertising rather than call-out “villains,” Jędrocha said. Also, it’s possible data providers start with accurate information and the mismatching happens elsewhere in the digital ad supply chain—and it could be hard to pinpoint the parties responsible for any breakdown. “The stats, they are a bit alarming,” said Jędrocha. “A lot of third-party data that is out there … it’s not fully transparent on how those data segments are built. We wanted to dive deeper into that.” Adlook, which is a part of Poland-based RTB House Group, considers contextual advertising—based on the content people consume, the sites they visit and their search habits—as a better way to build audiences relevant for brands rather than the pervasive data markets in the programmatic ecosystem. Its survey focused on U.S. internet users, and it tested the accuracy of audience segments provided by third-party data sources and activated through demand-side platforms (DSPs). Ad buyers on popular DSPs can pay for extra data from third-party sources to increase their odds of reaching a desired group. Adlook said it used widely available targeting segments. Some of the findings: 40% of people classified as “homeowners” were renters. When Adlook targeted “women 18 to 24,” only 18% of the impressions went to that cohort. In fact, 43% of the impressions went to men and 35% to people 55 and older, according to Adlook. One of the key findings revolved around targeting parents, 67% of whom turned out to not have children. The quality of data is a hot topic in digital advertising, where brands are trying to be as on-target as possible with their messaging. But consumer information, collected offline and online, can be low quality. Also, making use of data segments online is becoming more complicated as the ad tech industry operates under new privacy restrictions. Platforms such as Apple have eliminated third-party cookies on web browsers, which are an easy way to detect browsing habits and serve ads based on those expressed preferences. There also is a more fractured digital landscape with consumers switching among devices, from desktop to mobile to connected TVs, and across platforms. Brands are coming up with new strategies, such as collecting more first-party data, to identify key audiences for their products and then hitting them with online ads. Machine learning, artificial intelligence and new data and identity products are also cropping up to facilitate digital advertising. Ad Age’s guide to post-cookie advertising Brands and agencies use data management platforms, where first-party data collected from customers is combined with third-party data to build audiences for online campaigns. Targeting can be easier, and more accurate, within so-called walled gardens such as Meta, Amazon and Google, where there are logged-in users with known characteristics. But on the open web, which comprises the rest of the internet, it can be difficult for advertisers to hit their mark with programmatic advertising. Advertisers can buy audience segments from data providers that are available directly within the DSPs. “That data is not cheap,” Jędrocha said. “The third-party segments go anywhere from 15 cents, up to $2.50, or even more for some very niche audiences.” Adlook’s study was not the first to suggest that online identities can be quite elusive. Last year, Ad Age’s Jack Neff took a personal “data journey,” and found that some of the biggest brokers mischaracterized basic traits about his personal interests. One data broker said Neff lived in Little Rock, Arkansas, which he had never even visited. The erroneous profiling led to mismatched ad targeting in digital spaces—Neff had received ads in Spanish from Procter & Gamble even though he didn’t speak the language. The Adlook experiment points to the complexity in programmatic ad markets. The future of internet advertising, however, could lend to new ways to build consumer profiles instead of traditional demographic targeting. “When you are a skincare brand, you’re not really interested in ‘women 20-44 from big cities,’” Jędrocha said. “You want to target people interested in personal care or in-market for skincare products.”

Ad targeting failures—why brands are missing their intended audiences

Featured Stories