Raghu Kodige, CEO and co-founder of Anoki, breaks down one of the biggest problems plaguing the connected TV (CTV) ecosystem: lack of effective targeting. With no cookies and limited first-party data, CTV publishers are struggling to differentiate their ad inventory—leading to lower fill rates and CPMs. Kodige says this is exactly the issue Anoki set out to solve using multimodal AI.
Anoki’s technology identifies the scene-level context of a show or movie and feeds that data into the ad server in real time. If a viewer is watching a scene set in a bar, for example, the system flags that context so an advertiser can serve a beer ad rather than something irrelevant, like a baby product. The results are better engagement for advertisers and better monetization for publishers. The solution has gained major traction—Anoki recently announced integrations with Magnite and The Trade Desk’s new Ventura OS, signaling serious momentum from the buy and sell sides alike.
Raghu Kodige:
Yeah. Hi. I am Raghu Kodige. I am the CEO and co-founder of Anoki. And Anoki is a AI company in the CTV space. We use our multimodal AI to both improve the content discovery as well as improve monetization for various content partners that we work with. This is my third show, and I love that it is growing bigger and bigger each year. So the biggest problem in CTV for publishers and content partners is that they're not making enough revenue, which means either the fill rates are low or the CPMs are low, or it's a combination of both. And the main reason for that is unlike traditional web where there are cookies for advertisers to be able to target precisely, CTV does not have that, right? There is no cookies in CTV, which means there is no real first-party data, and so everyone's inventory looks the same and an advertiser is trying to buy it, and that causes CPMs and full rates to keep going down.
So Anoki has come up with a multimodal AI solution for understanding what is the context someone is in when they go into the ad break. And we do this at the scene level. So what I mean by that is, let's say you're watching a movie and it's set in a bar, then the best ad to serve to that consumer when it goes into the ad break is probably a beer commercial and not a baby product. So in order to inform that advertiser, we detect the scene-level context, in this case, the bar scene, and pass it into the ad call so that an advertiser can appropriately put the right ad in that place.
Yeah, so Anoki has been in the market with this solution for over nine months now, and the reception has been great. Most recently, just last week, Magnite which is one of the largest SSPs announced that they have fully integrated this solution into the ad server and SSP that Magnite has. So now this gives advertisers a large amount of scale to be able to buy contextually relevant advertising through Magnite. Just today we also announced our partnership with Trade Desk Ventura OS. So very soon our app and the technology will be part of the Ventura OS as well. So we are working with some of the largest companies in our space on the SSP and the DSP side to bring this technology to advertisers and help increase the revenue for everybody.