SAN JUAN, Puerto Rico — Companies that focus artificial intelligence implementation with an eye toward business growth, rather than expense reduction, will outpace competitors. The view is that AI’s transformative power should drive real outcomes that matter to brands rather than operational efficiency alone. But getting brands to accept that remains a challenge.
“Who’s going to leap ahead of their competition are brands and publishers who are focused on using AI for growth, not necessarily cost cutting,” Ali Manning, co-founder & COO of software platform Chalice told Beet.TV contributor David Kaplan at the Beet Retreat San Juan. “That’s something that will happen, but the AI focus should really be on how is this going to grow my brand?”
This approach transforms advertising markets into innovation engines where advertiser dollars become investments rather than expenses, rewarding quality publishers while enabling brand growth.
From proxy metrics to business outcomes
Leading brands move beyond platform-driven proxy metrics like clicks and video completion rates toward C-suite relevant outcomes, with companies like Hershey’s using AI to drive incremental store sales and Bayer modeling new-to-brand customers for Claritin at shelf level.
“Since we’ve been beholden to a few companies that have held the power of AI models for themselves, they’ve forced advertisers into buying proxy metrics,” Manning said. “They’re not really about the advertiser, the brand’s business. They’re about the platform’s ability to serve a bunch of advertisers at once.”
Successful implementations focus on outcomes executives discuss rather than marketing manager optimization requirements.
Custom models deliver enterprise value
Most companies handle one or two AI implementation components well, but maximum value requires integrating best available data, optimal models, and relevant outcome prediction, with custom enterprise models outperforming generic solutions.
“Most don’t predict an outcome that really matters to the brand. Most companies give every customer the same AI model, and that’s just not the way enterprises are going to get the most use out of AI,” Manning said. “You don’t want to use the same model if you’re one cell phone company as your two other competitors. You want a model that’s built and deployed directly for you.”
“Messy” data beats perfect delay
Brands should begin testing with available data rather than pursuing lengthy data cleaning projects, as Chalice works with advertisers using messy data and consumer packaged goods companies with minimal datasets.
“You don’t need all of the data in the world. You don’t need it to be clean. We work with advertisers who have messy data. We work with CPGs who have essentially none,” Manning said. “What we do is we find the best available data and start modeling on that. That’s so much better than trying to do a one year data project that turns into a five year data project.”
CTV signals create pricing opportunities
Connected TV presents AI opportunities beyond reach optimization through rational pricing based on consumer access frequency, with high-value consumers requiring different investment levels based on viewing availability.
“If I am a brand and I have a consumer who is very high value to me, that consumer might be someone who watches a lot of non-premium content, or they could be a consumer who you can only get once a week,” Manning said, citing working mothers with limited viewing windows.
Chalice partners with publishers like Paramount to access show-level data for enhanced decisioning capabilities, with select publishers positioned to transform market dynamics through the next upfront cycle and beyond.
“It’s going to be a few publishers who are going to really transform and run ahead of the market in this next upfront cycle,” Manning said.
You’re watching coverage from Beet Retreat San Juan 2026, presented by Alliant and TransUnion. For more videos from this series, please visit this page.
