If AI Buys Everything on Your Behalf, Where Does Advertising Fit In?

If AI Buys Everything on Your Behalf, Where Does Advertising Fit In?

Todd and Kelly dive into a provocative Harvard Business Review article exploring how AI-powered commerce agents in China are already making purchasing decisions on behalf of consumers — and what that means for the entire marketing industry. Todd argues that these agents operate purely on algorithms, structured data, and utility signals, bypassing brand affinity and emotional connection entirely. Kelly pushes back, insisting that human preferences, emotional buying behavior, and brand loyalty still play a critical role — especially in US markets where consumers aren't ready to hand over full purchasing autonomy to a machine. The two debate whether performance marketing, attribution models, and traditional advertising funnels can survive in a world where the "customer" is an algorithm, not a person. They explore Google's Zero Moment of Truth framework, draw parallels to self-driving cars, and land on a shared conclusion: upstream strategic thinking and creative problem-solving will be more valuable than ever for marketing leaders navigating this shift. Chapters: * (00:00:00) - Cold open: How advertising shapes buying decisions today * (00:01:05) - Episode intro and Arsenal heartbreak * (00:02:00) - The HBR article: AI agents are conducting commerce in China * (00:05:05) - What happens to performance marketing when agents buy for you? * (00:07:10) - Todd vs. Kelly: Do AI agents follow brand preferences or pure algorithms? * (00:11:00) - Todd's case: Agents optimize on price, utility, and data — not loyalty * (00:14:00) - Kelly's counterpoint: Humans will never fully surrender emotional buying * (00:18:30) - Attribution in a black box: Can you track what an AI agent does? * (00:24:00) - The Zero Moment of Truth and why upstream marketing still matters * (00:31:00) - Final take: What CMOs and agencies need to prepare for now Links and Resources: * Kelly Maguire on LinkedIn [https://www.linkedin.com/in/kmaguire/] * Todd Juneau on LinkedIn [https://www.linkedin.com/in/toddjuneau/] * Vuja Dé Digital [https://vujadedigital.com/] Thanks so much for joining us this week. Want to subscribe to Contrary to Popular Opinion? Have some feedback you’d like to share? Connect with us on Spotify [https://open.spotify.com/show/7zfdNmA8zehTU9RCjeYSzI], Apple Podcasts [https://podcasts.apple.com/us/podcast/contrary-to-popular-opinion/id1743279046] and YouTube [https://www.youtube.com/channel/UCzxu33-b3R9QV3M1HJJvITQ] to leave us a review!

A 'Proven Agency Process' is Only Valuable if it Fits Your Business

Todd and Kelly address the real trade-offs between bespoke, tool-agnostic agency work and “proven framework” agencies, and when bringing marketing agency work in-house actually helps or hurts. They debate why productized agency “machines” sell so well (trophies, repeatability, and perceived certainty), why truly custom solutions are operationally harder but often better aligned to business realities, and how marketer expectations can get misaligned when a brand wants F1 performance with NASCAR budgets. They also dig into the talent, incentives, and learning dynamics that make it difficult for most brands to keep a truly top-tier media and marketing function in-house, unless the brand has enough scale, specialization, and leadership to sustain it. Chapters: * (00:00:00) Bespoke vs framework agencies: the core tension * (00:03:10) Tool-agnostic strategy: why “fit” beats defaulting to one platform * (00:06:20) Incentives, closed ecosystems, and where hidden money can show up * (00:08:50) Why the “shiny machine” sells (and why bespoke is harder to buy) * (00:11:20) Are clients paying for learning curves—or for edge? * (00:14:20) The F1 car analogy: trophies vs building the right car for the race * (00:17:00) Matching the “race” to the business: maturity, budgets, and constraints * (00:21:50) In-house vs agency: the talent and learning opportunity problem * (00:26:30) When in-house can work: scale, specialization, relationships, leadership * (00:30:10) The concentration risk: what happens when key people leave Links and Resources: * Kelly Maguire on LinkedIn [https://www.linkedin.com/in/kmaguire/] * Todd Juneau on LinkedIn [https://www.linkedin.com/in/toddjuneau/] * Vuja Dé Digital [https://vujadedigital.com/] Thanks so much for joining us this week. Want to subscribe to Contrary to Popular Opinion? Have some feedback you’d like to share? Connect with us on Spotify [https://open.spotify.com/show/7zfdNmA8zehTU9RCjeYSzI], Apple Podcasts [https://podcasts.apple.com/us/podcast/contrary-to-popular-opinion/id1743279046] and YouTube [https://www.youtube.com/channel/UCzxu33-b3R9QV3M1HJJvITQ] to leave us a review!

The Real Cost of AI: What Most Agencies Won't Tell You

Todd and Kelly address a growing tension in marketing: expectations that agency work should cost less because of AI, even as the cost of senior talent and modern tech stacks continues to rise. They argue that AI is powerful at improving the execution layer, including generating options, accelerating testing, and enhancing optimization workflows, but it cannot own judgment or decide what truly matters to a business. The conversation breaks down why “process agencies” often rely on junior teams and automation, why that model can amplify bad strategy, and why long-term success still depends on experienced marketers, transparent measurement, and aligned incentives that reward partners for going beyond “good enough.” Chapters: * (00:00:00) Intro: AI, agency economics, and why this topic matters now * (00:01:00) Inside the indie agency retreat: what the panel got right and missed * (00:02:30) The question clients avoid: why talent + tech is not getting cheaper * (00:04:00) What AI is actually improving: the execution layer and pitch theater * (00:05:30) The risk: junior teams + automation turning agencies into “process shops” * (00:06:45) Strategy still needs people: judgment, experience, and what matters most * (00:08:00) What clients are really buying when they hire an agency * (00:10:00) Short-term thinking, CMO pressure, and the “throw money at it” trap * (00:11:15) AI incentives and laziness: why cheap engagements lead to bare minimum * (00:13:10) Why aligned incentives matter: performance upside and shared wins * (00:16:00) How AI works (and doesn’t): probabilities, not reasoning or business context * (00:19:30) The team you actually need: strategy, account leadership, specialists, analysts * (00:25:00) The core contradiction: higher costs, higher expectations, lower fees * (00:28:30) Why clients choose the shiny pitch and then feel buyer’s remorse * (00:32:00) Procurement “guarantees” and why handcuffs break performance marketing * (00:36:30) The hidden risks: narrative control, papering over cracks, and trust * (00:38:15) “Rounding error” clients: why top talent moves to bigger, better-fit accounts * (00:40:15) The better bet: pay fairly, get priority, and use AI to raise outcomes * (00:41:45) Closing challenge: CMOs, explain why you expect more for less Links and Resources: * Kelly Maguire on LinkedIn [https://www.linkedin.com/in/kmaguire/] * Todd Juneau on LinkedIn [https://www.linkedin.com/in/toddjuneau/] * Vuja Dé Digital [https://vujadedigital.com/] Thanks so much for joining us this week. Want to subscribe to Contrary to Popular Opinion? Have some feedback you’d like to share? Connect with us on Spotify [https://open.spotify.com/show/7zfdNmA8zehTU9RCjeYSzI], Apple Podcasts [https://podcasts.apple.com/us/podcast/contrary-to-popular-opinion/id1743279046] and YouTube [https://www.youtube.com/channel/UCzxu33-b3R9QV3M1HJJvITQ] to leave us a review!

The Right Questions About Agentic Media Buying No One Is Asking

Todd and Kelly debate a widely discussed case study claiming an “agentic media buy” bypassed the DSP to reduce supply-chain costs and shift more budget into working media. They address what DSPs actually contribute, why headlines about “cheaper and faster” can be misleading, and how cost savings often reappear elsewhere in the ecosystem. The conversation moves from the mechanics of media buying to the human realities of capacity, burnout, and talent costs. They ultimately frame agentic workflows as an evolution that may remove entry-level tasks while increasing the premium on senior talent, strategy, and decision-making. Chapters: * (00:00:00) Opening and why “agentic media buying” is getting hype * (00:02:30) The case study claim: bypassing the DSP and reallocating fees to working media * (00:05:20) What a DSP actually does and why it's not “inherently bad” * (00:08:10) Why “savings” rarely stay savings: value shifts and someone monetizes it * (00:11:40) Agency economics: fee pressure, utilization, and the race to the bottom * (00:16:00) The human constraint: focus, burnout, and “AI brain fry” * (00:20:10) What agentic systems may automate (setup, trafficking) vs. what stays human (judgment) * (00:24:40) The risk of paying less and getting worse, and why “better outcomes” should be the goal * (00:31:50) Where this goes next: evolution, talent getting more expensive, and the entry-level gap Links and Resources: * Kelly Maguire on LinkedIn [https://www.linkedin.com/in/kmaguire/] * Todd Juneau on LinkedIn [https://www.linkedin.com/in/toddjuneau/] * Vuja Dé Digital [https://vujadedigital.com/] Thanks so much for joining us this week. Want to subscribe to Contrary to Popular Opinion? Have some feedback you’d like to share? Connect with us on Spotify [https://open.spotify.com/show/7zfdNmA8zehTU9RCjeYSzI], Apple Podcasts [https://podcasts.apple.com/us/podcast/contrary-to-popular-opinion/id1743279046] and YouTube [https://www.youtube.com/channel/UCzxu33-b3R9QV3M1HJJvITQ] to leave us a review!

Cookies Aren't Dead. They Just Evolved.

Todd and Kelly break down the practical difference between zero-party data (what people willingly tell you) and first-party data (what you observe from owned interactions), and why the industry’s privacy shift has not eliminated tracking so much as it has made measurement more fragmented and harder to interpret. They debate whether cookies will ever truly go away, why consumer behavior tends to default to convenience (even when privacy is a concern), and how “sleeping giants” with money and incentives will keep finding new ways to build audiences. From there, they explore where synthetic personas fit in today: not as a magic replacement for real data, but as a way to pressure-test messaging, creative, and experiences before spending heavily. Finally, they zoom out on the bigger measurement challenge: when direct attribution gets noisy, the next frontier is faster, more accessible pattern recognition and media mix-style modeling powered by AI, which can help marketers ask better questions, explore more hypotheses, and make more confident decisions across long, multi-touch journeys. Chapters: * (00:00:00) Cold open: why data gets collected “one way or another” * (00:01:10) Intro banter, then diving into zero-party vs first-party data * (00:02:00) What is zero-party data, and where does it come from? * (00:05:10) Why it matters most for long decision journeys (healthcare, high-consideration) * (00:07:15) The real marketing challenge: multi-stage investment and ROI math * (00:09:00) Tracking, privacy, and why “accept all” is still the default behavior * (00:14:20) Synthetic personas: where it helps today (creative and message testing) * (00:20:40) Is targeting better or worse than 5–10 years ago? * (00:23:20) Why AI changes analytics: faster hypothesis testing and pattern discovery * (00:34:40) The risk: AI can scale bad assumptions as fast as good ones Links and Resources: * Kelly Maguire on LinkedIn [https://www.linkedin.com/in/kmaguire/] * Todd Juneau on LinkedIn [https://www.linkedin.com/in/toddjuneau/] * Vuja Dé Digital [https://vujadedigital.com/] Thanks so much for joining us this week. Want to subscribe to Contrary to Popular Opinion? Have some feedback you’d like to share? Connect with us on Spotify [https://open.spotify.com/show/7zfdNmA8zehTU9RCjeYSzI], Apple Podcasts [https://podcasts.apple.com/us/podcast/contrary-to-popular-opinion/id1743279046] and YouTube [https://www.youtube.com/channel/UCzxu33-b3R9QV3M1HJJvITQ] to leave us a review!