Categorizing Attention Tracking Partners: What Marketers Need to Know About This Ad Metric

The ad industry is now brimming with vendors claiming they can track and measure ad viewers' attention. Hungry to strike partnerships with agencies and their clients, these vendors want to prove that running engaging ad campaigns is not just an art, but a science. The trouble is, attention-tracking and measurement vendors are still developing their (mostly unvetted) offerings. Because the attention economy is so young, most media buyers aren’t yet familiar with their partnership options in the category—let alone up to speed on how one vendor is different from another. However, 90% of respondents to a 2022 Realeyes survey of 320 ad industry responders said attention metrics are important. Over half of the group indicated their company would invest in attention metrics within the next 12 months. May the Best Attention Metric Win Without a standardized attention metric or any industry alignment on which factors reliably predict attention, vendors' approaches to tracking and measurement aren't all the same. As advertisers parse offerings and seek the solutions that lend the most lift to clients' campaigns, they'll determine which methodologies work best, and which don't support their desired results. "You've got dozens of companies that have come out with 'attention metrics.' Everybody's calling it the same thing. [There are] lots of different companies that are approaching things in slightly different ways," Max Kalehoff, Realeyes vp of growth and marketing, told Adweek. Vendors fall in two categories There are scant industry resources illuminating what attention firms do, but a few credible guides exist. Realeyes offers an Attention Technology Providers Directory on its website, and Kalehoff recently published a market research report on WARC. Kalehoff's report and interviews Adweek conducted with vendor leaders revealed most partners fall into one of two categories: Human-based vendors: Firms that collect visual data using consenting human subjects. They use tools like front-facing cameras to track eye movement and draw conclusions about which media or creative elements most engage viewers. Sometimes, these vendors use AI or machine learning to draw conclusions from the data they gather. They may also license their data and technology to other vendors, agencies and marketers.Machine-based vendors: Firms that use algorithms or AI to determine which planning conditions, like the channel type or viewing medium, are most likely to engage viewers. To inform their predictions, these vendors sometimes license eye-tracking data from human-based vendors. Adweek interviewed leaders at Adelaide and Amplified Intelligence—two vendors that take different approaches to attention research— to learn more about the two categories they represent. Machine-based vendors We start to train the algorithm so the [attention unit] that it produces is as close to a proxy for outcomes as possible.–Marc Guldimann, founder and CEO, Adelaide Adelaide More vendors in this category: DoubleVerify, IAS, PlaygroundXYZ and Moat Attention measurement firm Adelaide relies on its proprietary algorithm to gauge the probability that viewers pay attention. It created the attention economy's most widely recognized metric—a probability score between 0 and 100, called an Attention Unit (AU). It's "a global, omnichannel, apples-to-apples metric that predicts the likelihood of attention, but also the likelihood of subsequent outcomes," Adelaide CEO Marc Guldimann told Adweek. Guldimann’s clients can use the AU regardless of if they're buying, planning or simply measuring impressions. "We've benefited from some of our dogma around creating just one metric, and that metric being a probability score," Guldimann said. Adelaide built its machine-learning algorithm using eye-tracking data it licensed from human-based vendors. To generate the AU score, the AI considers elements like how large an ad is relative to the rest of the page, how long the ad is on the screen, how quickly the page moves, audibility, the ad’s position on the page relative to the center, and how many other ads are on the page. The technology also analyzes impressions gleaned from cookie data to strengthen its predictions, as well as how its clients' attention-optimized campaigns actually perform. "We start to train the algorithm so the AU that it produces is as close to a proxy for outcomes as possible," said Guldimann. Solutions like Adelaide’s represent “serious opportunity for comprehensiveness, scale and real-time campaign optimization,” according to Kalehoff's report. But advertisers are still debating whether or not they should consider campaign lift gleaned from impression proxies as an indication of increased consumer attention. Inside Dentsu Media’s 5-Year Endeavor to Measure Attention Human-based vendors Filming humans en masse, to the scale of traditional reach panels, has cost and ethical implications.—Karen Nelson-Field, founder and CEO, Amplified Intelligence Amplified Intelligence More vendors in this category: Lumen, Realeyes and TVision Amplified Intelligence's offering is human-based and reliant on what its founder and CEO, Karen Nelson-Field, calls "person-level information." The firm runs observed human panels, with a 35,000-large sample set. It gathers facial data from ad viewers in real-time, with their consent. Its scientific approach gives it an edge on competitors that don't have the access, or the ability, to conduct similarly extensive research. It is an ideal partner for agencies that wish to commission custom research studies and leverage findings to inform campaign planning. "Filming humans en masse, to the scale of traditional reach panels, has cost and ethical implications. No one really wants to have a currency that is, by default, surveillance," Nelson-Field said. "And I understand that, but it's highly, highly accurate." Amplified Intelligence's research capabilities and ability to gather human data make it an attractive partner for agencies that want to co-invest in new research, or feed the firm's existing data into their own tech. Omnicom agency OMD built its proprietary Attention Requirement Calculator with data gleaned from its Amplified Intelligence research partnership. How OMD’s Attention Planning Tool Transforms Campaign Strategy Amplified Intelligence's relationship with OMD is one example of how agencies can leverage partners' eye-tracking data and research to build proprietary hybrid solutions. While agencies partner closely with one or more vendors in the space, there is a tendency toward bias. Agency leaders' opinions about which tracking firms are most effective "often tend to correlate a bit too conveniently and too closely to their own business model, and what kind of tools they've got," Scott McDonald, president and CEO of the Advertising Research Federation, told Adweek. Agencies and marketers alike must now determine which partners and methodologies are reliable, and which vendors will emerge as industry leaders in the long term. "Nothing is in a final, perfect state," said Kalehoff. "The way you advance the art and the science is to have smart and often competing ideas."