For advertisers itching to use automatic content recognition (ACR) to target their messages to gamers on their CTVs, 2023 may be their year.
While TV manufacturers and data brokers have been compiling ACR data for years now, it’s still early days when it comes to using ACR signals derived from gaming for actual ad targeting. But the expected expansion of in-game advertising to console games this year should unlock more applications of ACR data in gaming.
Thanks to ACR, CTV devices can pinpoint the specific games being played and how far a gamer has progressed through a game. Are they winning or losing? ACR data can figure it out. CTVs can also tell what external hardware is connected to their HDMI ports, which means they know what gaming consoles their owners play.
All these data points, when shared with marketers, can make them think more strategically about what messaging will resonate with gamers. This data can also be used to create targeting signals that can be ingested by DSPs. But most DSPs are in the early stages of integrating these signals into their platforms.
ACR in gaming
ACR technology determines the exact content being displayed on the screen by comparing the image against a library of reference images, sometimes referred to as “fingerprints.”
If a gamer is opted into ACR, their CTV device automatically captures data on what video games they’re playing, as long as they’re part of an ACR tech’s reference library. (Users typically opt into allowing their onscreen content to be captured by ACR when they initially set up their TVs.)
ACR capabilities built into the CTV device itself can take snapshots of the onscreen content as often as every 500 milliseconds, said Corey Johnston, Samsung Ads’ sales manager for Canada. That means TV manufacturers capture first-party data on where gamers are in their gameplay journey in close to real time.
For example, in console games with story modes that follow a predetermined level progression, ACR can recognize specific levels or specific cutscenes to determine how much of a game a player has completed. “Game publishers will ask us to fingerprint a specific area of their game to understand if someone gets there or not,” Johnston said.
While ACR is capable of recognizing granular signals, like specific scenes from a video game or imagery captured during gameplay, the technology can also interpret more obvious signals, like a game’s startup screen or pause menu. These signals are often a simpler, more reliable way to determine which title a gamer is playing than interpreting scenes from the actual gameplay.
ACR and targeting
Game developers can use ACR signals for a variety of marketing use cases. For example, if ACR data determines that a player has completely beaten a game’s story mode, they could be targeted with ads for downloadable content that introduces new levels or gameplay experiences to the game.
For brands that already advertise in video games, ACR can recognize when their ads appear onscreen during gameplay. Then gamers can be retargeted based on that exposure.
This kind of retargeting is already happening in streaming and linear TV. If an ACR provider puts brand logos in its reference library, brands could buy an ad in the pod immediately after the brand appears, according to a Samba TV spokesperson. For example, if someone drinks a Coke in an episode of “Yellowstone,” a soft drink manufacturer may want to buy a spot in the ad pod after the character quenches their thirst.
Likewise, if a gamer is playing an NBA 2K game and participates in the Mountain Dew 3-Point Contest, ACR can recognize the Mountain Dew branding on the court, and that gamer could then be retargeted with Mountain Dew ads.
ACR signals also allow marketers to tailor their ad messages according to how well a gamer is performing at a given time. For example, if ACR picks up on signals like a “Game Over” screen appearing multiple times in a short time frame, an ad could be targeted to that player for a power-up they can purchase from the game developer. Or they could be served an ad for a Twitch stream that offers gameplay tips for that specific game.
Of course, most of these use cases are currently hypothetical.
One area where gaming ACR data will be immediately actionable is to reach a video game’s preexisting fan base in order to attract viewers to TV and movie adaptations of those games.
For example, HBO’s adaptation of Naughty Dog’s “The Last of Us” video game will debut this month. “The Last of Us” and its sequel have sold more than 37 million copies worldwide, so it stands to reason that HBO wants to tap into this massive prebuilt audience to promote the new series.
What’s more, an updated version of the original 2013 game was released in September 2022. So ACR data can facilitate targeting any “The Last of Us” superfans who have been replaying the game in anticipation of the series premiere.
“If someone has played these games, these are key, highly engaged segments for advertisers trying to drive viewers to their show,” Johnston said. “It’s a no-brainer audience for them.”
Targeting players of a specific game on CTV, like for an adaptation of “The Last of Us,” is currently the most mature use case for ACR data. And players can be bucketed into custom audiences for cross-channel targeting based on their favorite games and genres. But the next opportunity will be using that data to target ads within the games themselves.
Correction 1/9/2023: A previous version of this article quoted Samsung Ads’ Corey Johnston as saying ACR can capture onscreen content as often as every five seconds. ACR can actually capture onscreen content as often as every 500 milliseconds.
