As fragmentation across video services has obscured visibility into what consumers are watching on TV, ACR (automated content recognition) data has emerged as a valuable tool for buyers looking to reach audiences across CTV. A recent survey from LG Ad Solutions found that more than half of CTV buyers use ACR data for targeting, and consider it “very valuable” when planning video campaigns.
But the role of ACR data is evolving, from its more standard application in linear TV reach extension, into activation and measurement use cases. Yet there remain limitations in using ACR data, particularly when it comes to measurement, and the blocking of data collection by big players in the CTV ecosystem.
Chasing lost reach
ACR technology began to be embedded in smart TVs more than 10 years ago, in order to detect and identify the content that is being watched in real-time. Karim Rayes, Chief Product Officer at Nexxen, says this ‘on the glass’ approach offers wider capabilities than the panel-based data collected by the likes of Nielsen or Barb.
“The benefit of ACR versus a panel is that you’re really on the glass, so you can capture everything, as long as you have user consent, whether it’s coming through HDMI, through the apps, or through linear TV,” comments Rayes. “This means you can go beyond movies and TV. We’re starting to explore gaming as an example.”
But first and foremost the technology became useful for planning across linear TV and streaming services. While panel-based measurement provided a view of reach for linear TV spots, the fragmentation of viewing caused those audiences to decline, leaving TV buyers looking to rebuild that reach in digital environments. ACR data helped them find audiences across channels and streaming services, using a consolidated view of CTV viewing to mix digital buys into TV campaigns. The reverse also applies, where digital-only buyers could use the data to bring linear TV into the mix.
And reach extension remains the primary use case for AV planners, according to Tom Stimpfig, Head of Video at PHD. Using ACR data from Samba TV, the agency can identify which postcodes have been underexposed to its own linear TV, BVOD or CTV campaigns, and retarget those audiences in other channels. The data is also used for competitor retargeting, by identifying which postcodes have been exposed to ads from competitor brands across linear TV and streaming services, then countering that message accordingly, or simply sharing that insight back to the clients.
“Our first-party ACR data can pinpoint postcode areas where ad exposure is over- or under-delivered, helping advertisers optimise linear TV spend, reduce wastage, and target under-reached audiences more efficiently,” comments Dan Conneally, VP, Agency Solutions UK at Samba TV. “These geotargeting insights can also inform planning across other channels, including social, DOOH, and radio, providing a holistic view of audience reach.”
The ACR misconception
But the use of ACR data now goes beyond planning and retargeting, says Conneally. “While that was an important early use case, today ACR fuels a much wider set of applications, from cross-platform measurement and attribution to competitive benchmarking and audience intelligence,” he comments. “The misconception is to think of ACR as a narrow tool, when in fact it has become a foundational layer of media intelligence, shaping how advertisers plan, activate, and measure campaigns with confidence.”
On the activation front, audiences derived from ACR data can be plugged into demand-side platforms (DSP) and supply-side platforms (SSP), enabling activation teams to run programmatic campaigns by transacting on deal IDs. Nexxen’s Karim Rayes says that directly activating plans built on ACR data streamlines the programmatic process. “You’ve planned on data, now you activate on the same data.”
PHD’s Tom Stimpfig notes that ACR vendors are increasingly simplifying that activation piece by creating partnerships with DSPs, such as The Trade Desk or Google DV360. This allows PHD to push its target postcodes into a DSP and activate against those audiences. “We can also use the postcode data to activate across BVOD services, which might not be fully integrated into a DSP in the same way that some other CTV partners are,” he adds. “I would say most of our advertisers, if they’re running on linear TV, use some form of ACR data within either their planning or their activation.”
Missing pieces
ACR technology also serves a growing number of measurement functions, such as measuring incremental reach by seeing how many households were watching an ad spot, or tracking tune-in lift for entertainment companies that want to see how many viewers tuned into a new show after seeing it advertised. And by onboarding sales data and web signals, ACR vendors can also measure in-store purchases, website visits or game downloads. “You’re essentially able to tie your digital campaign to the data we see from the TV,” says Rayes.
However, Stimpfig observes that the measurement functionality of ACR data is limited, and more work is required before it is widely adopted as a measurement tool. ACR data covers household rather than individual viewing, meaning it cannot be matched back to audience data provided by joint industry currencies (JICs) such as Barb.
“This remains a challenge for trying to use ACR data for incremental reach measurement,” says Stimpfig. “In five years time, it may look slightly different. There are a lot of people doing incremental reach and frequency management in C-Flight, Origin and AudienceProject, but they are all largely able to integrate with Barb in some way, whereas ACR currently is not.”
Certain streaming companies (most notably Netflix) will also block ACR vendors from being able to collect data from their apps, athough Rayes notes that these are in the minority. “You have a couple of the bigger folks that do block, so you do have a blind spot when it comes to these,” he says. “But when it comes to the ad-supported apps, such as Pluto TV and so forth, those are less restrictive.”
And PHD has been able to use ACR data to track the performance of campaigns while they are live, in order to optimise them and demonstrate that value back to clients. Stimpfig says that when its campaign went live, the agency used the first tranche of data from Samba TV to see the 2,000 postcodes that fell into the bottom quartile in terms of reach exposure. By targeting these postcodes each week, only 500 of them remained in the bottom quartile in the final run of data.
“While that’s not reach and frequency measurement, it does give us a good indication that ACR data is useful in trying to act as a reach extender,” he remarks. “And clients find that quite useful, especially when you’re able to map it back and say, these were all the postcodes that were initially underexposed and over time this is how it’s reduced on a UK map. So we are able to show its value.”
Follow VideoWeek on LinkedIn.
