• by Laurie Sullivan  @lauriesullivan, November 1, 2022

Digital intelligence company Similarweb has announced a collaboration with EDO, a TV outcome measurement company for convergent TV, that gives brands a way to use search data to measure the performance of television advertising.

“We believe syndicated data is very powerful,” said EDO President and CEO Kevin Krim. “The problem is that outside of the people who care about measurement, no one knows what that means.”

Krim said he and others fell into this method while doing work on how to get early indicators of changes in economic behavior and market share and sales. Web search and patterns in web traffic are good early indicators, he said. He said the co-founders realized it’s a way to measure ad efficacy.

“You get this in Google Trends, which was an early way to do this,” he said. “Search leaves huge anonymous breadcrumb trails.”

EDO will receive internet activity data from Similarweb hourly and match and measure lift in digital behaviors with the more than half a billion impressions per day that EDO observes across all of linear and streaming TV.

“The search and website data identifies where or not it changed hearts and minds,” he said. “We can see web searches and website visitors and app use and downloads, and retail football traffic. We do this through partners and through our own data.”

EDO scores and measures the performance for each TV ad based on lift and engagement by combining data from search and website visits. The measurement product is sold through EDO’s cloud-based software to brands and agencies, and media companies who own the networks.

The insights learned from this collaboration aims to help brands, agencies, and networks know the solutions work and how to get the most from consumer engagement. It will also expand EDO’s existing consumer engagement dataset for over the top (OTT), increasing the number of surveyed household panels from the hundreds of thousands to the millions. 

When an ad is shown to a different person each time, it creates an entirely new challenge, Krim said, especially with privacy regulations.

“You have all this signal loss from privacy regulations, so what do you so,” he said. “Marketers need to resort to other tactics like we do, take small chunks of huge panels in a certain ZIP code, and look at who was exposed to an ad campaign on average. On a group level, a cohort.”

From there analysts can see when one group was more likely to go to the website of the brand after seeing the ad, he said.

“The technique is called differential privacy or, more specifically, micro cohorts,” he said. “As far as we know, no one has used it before in this way.”