This article was produced in partnership with Viant
AI’s impact on the digital advertising ecosystem continues to be a significant story in 2026, and adtech firm Viant hopes to capitalize on this trend with its new product Outcomes.
Viant already has an automation system called Viant AI that was announced in late 2024 and that powers Viant’s ad buying platform. Outcomes is a product built on that platform, and is designed to run ad campaigns to achieve specific business goals, with minimal human involvement.
On release, advertisers can use Outcomes to meet campaign goals based on cost-per-action and return on ad spend. More features are in development to hit mid- and upper-funnel goals.
This solution was necessary to reckon with the digital ecosystems’ growing fragmentation said Viant cofounder and COO Chris Vanderhook.
“The complexity across six or seven channels, all the different formats, all the audience segments — it’s incredible,” he said. “Advertisers were telling us, ‘Hey, we need more automation. This is becoming too complex.’ That was the call we were answering.”
MacKenzie-Childs, which manufactures whimsical ceramics and dinnerware, participated in a beta test that pitted Viant AI against a human trader managing a campaign in a DSP. The human also had an advantage: they could use retargeting, where Viant could not.
But Viant’s solution beat the human trader by 2.3x on a straight cost-per-action basis.
“The media plan out of the gate was so relevant that Viant AI computed on the open internet the top placements of ads that should produce results, and it was right,” said Viant cofounder and CEO Tim Vanderhook. “And when it wasn’t right, it was making changes to pricing, it shut publishers off, and it rotated out creatives that stopped performing in a seamless, autonomous way.”
Viant also hopes Outcomes strengthens its competitiveness against walled gardens like Google, Meta, and Amazon, which dominate ad spend. The adtech firm is doing this by making Outcomes as transparent as possible.
“With Outcomes, you see where the media is going to run, you see what the performance is, you see the channel mix,” Chris said. “We don’t think we need to trade autonomy for transparency.”
Viant leverages its proprietary models and data to train Viant AI. One model was for supply quality, ranking ad placement in terms of high-quality traffic or low-quality traffic. Another focused on connected TV, combining audience data from IRIS.TV, which Viant acquired in late 2024, and Viant’s own Household ID, which allows for precise targeting in a post-cookie world.
“We’ve been in this space a long time. It’s incredible to have a clear line of how we’re going to operate the business on behalf of our clients only,” Chris said. “It makes it really easy for everyone who works at Viant to know what our mission is every day.”
