“The Sell Sider” is a column written by the sell side of the digital media community.
As more inventory becomes available programmatically in CTV, the issue of ad quality is coming up more often.
Publishers, concerned about the messaging in ads being delivered, increasingly rely on DSPs and SSPs to filter out ad content that may violate policies on lingerie, tobacco or gambling ads. A failure to screen those violations could lead to those platforms losing, or having limited access to, CTV inventory.
To address this challenge, The Media Trust recently introduced an AI-powered video ad categorization tool that analyzes an ad’s creative against more than 15 sensitive or highly regulated categories – ranging from adult content, violence and hate speech to pharmaceuticals, politics and gambling – to determine if the content is objectionable before it is shown.
“A couple of partners right now are really concerned with gambling and marijuana because of all the new legislation around it,” said Cory Schnurr, head of marketplace innovation at ad quality platform The Media Trust. “As those mature and states allow more of this, there are big brands that want to get involved in that – but there’s rules around where and how they can do that.”
The company works with DSPs, SSPs and exchanges to flag those ads before they’re delivered to downstream video channels where, unlike in mobile, it would be hard for a family in a household to scroll past an objectionable ad during, say, a Sunday morning news or children’s program.
The categorization tool is currently in beta with three major platforms, though Schnurr declined to name them.
For video SSPs, the tool removes a number of manual processes, such as reviewing an advertiser’s self-declaration in open real-time bidding requests based on the IAB Content Taxonomy – which advertisers use for brand safety strategies – and enforcing agreements with DSPs to not send certain ads.
DSPs and SSPs can upload their creative – images, video, native tags and more – through The Media Trust’s API in order to determine if an ad included, say, violence or vaping. The tool also pinpoints nuances, such as a person drinking alcohol in a travel ad or a movie trailer featuring guns and smoking.
Attention to CTV ad quality has increased along with the new money flowing to the channel. With $8.7 billion in programmatic ad spend expected to flow into CTV by 2022, according to eMarketer, video content providers are looking to block ads that might harm the user experience and negatively impact revenue.
The incentives are in place to address quality issues in CTV. SSPs want to ensure they’re protecting their suppliers by not delivering harmful ads, while DSPs want to maintain their supply relationships. Typically, those platforms are working with their upstream partners to make sure they’re compliant. Most of The Media Trust’s partners also work with ad verification providers like DoubleVerify and Integral Ad Science.
“Our AI is able to put these labels on the ads and then we send that back to our ad platforms … and they can choose what they want to do with it,” Schnurr said.
The tool also helps platforms navigate different policies publishers may have. For example, one publisher may be OK showing ads for CBD products but not THC, while another prefers the latter because it’s more relevant to their audience. Or a major news publisher may have rules against running ads for cryptocurrency, which an SSP may not be able to determine.
“It basically allows [the platforms] to control what kind of tolerance level they want for certain categories,” Schnurr said.