Kristina Kaganer on Why the Best Marketing AI Won't Be Built by Engineers

July 8
38 mins

Episode Description

Host Pesach Lattin sits down with Kristina Kaganer, CPTO at Navigator and founder of Wayfable, for a conversation about AI, profitability, and what fifteen years inside the big holding companies actually teaches you. ABOUT THE GUEST Kristina Kaganer is Chief Product & Technology Officer at Navigator, where she leads product, engineering, paid media, and ad operations to build a commerce media network. She is the founder of Wayfable, a personalized children's storytelling product she designed and shipped in two weeks using the Yoto developer API and ElevenLabs voice synthesis. Before Navigator she spent roughly fifteen years buying and building marketing technology at [Coty], [AB InBev], and Publicis Commerce, where she led an NLP tool to 210% revenue growth and delivered a COVID-era marketplace in three weeks. She holds a background in psychology, learned to code in 2018, and has built three MCP (Model Context Protocol) servers. She moderates a 1,800-parent community in Southeast London and speaks Russian and Ukrainian. ABOUT THE HOST Pesach Lattin is a veteran digital advertising journalist and the host of the ADOTAT Show, covering the people and arguments shaping marketing, adtech, and AI. He has spent two decades reporting on and inside the advertising technology industry. WHAT WE COVER Why Kaganer argues AI is about access to knowledge, not automation, and the one question that kills most automation projects Her thesis that the best marketing AI gets built by psychologists and behavioral scientists, not computer scientists ("It's 100% of their job, 2% of someone else's") The "Cut the Crap" framework (Can't Realize Any Profit) and whether the e-commerce industry actually listened five years on ROAS vs. incrementality, retail media fragmentation, and last-mile delivery platforms The Publicis / Trade Desk transparency dispute and what it really signals about margin The real state of consumer AI adoption versus the hype, and the gender gap in usage How a toddler's toy, the Yoto developer API, and vibe coding produced a shipped product in two weeks Building your own support infrastructure as a woman in tech, against the backdrop of 40,000–60,000 women leaving UK tech roles annually Why chemistry is a forcing function for speed and quality KEY TERMS DISCUSSED Marketing technology, retail media, commerce media network, incrementality, ROAS, clean rooms, composable / vendor data stack, MCP servers, generative AI, vibe coding, ed-tech personalization, profitability vs. top-line revenue. Pesach Lattin / ADOTAT / adotatcom

See all episodes