AI agents can change how people discover products, compare brands, and decide what to buy. That raises a practical question for marketers: what should a brand do when software helps make the choice?
In this conversation, Mike Mallazzo, general manager of Alephic Media, makes the case for better product information, a consistent brand, and shopping experiences people want to spend time in. He also discusses his early experience with ChatGPT ads and where advertising aimed at agents might go.
Four takeaways
Make products easy to understand. Rich descriptions and consistent information help people and agents understand what a product does and who it is for. Mike sees this as useful marketing work whatever the pace of adoption.
Build demand for the brand itself. A customer who asks an agent for a specific brand gives it a different job from one asked to find an unspecified product. Mike argues that brand preference remains valuable as more decisions involve AI.
Use faster experimentation while it offers an advantage. Mike expects AI to lower the cost of producing and testing ads. He sees an opportunity for early adopters, while predicting that many of those advantages will narrow as the tools become common.
Keep discovery enjoyable. Agents can remove repetitive work and introduce people to products they would otherwise miss. Mike argues that browsing, anticipation, and choosing are also parts of shopping that people value.
In this episode
00:00 Meet Mike Mallazzo
An introduction to Mike and the question running through the episode: how marketing changes when agents help people make purchasing decisions.
00:35 What agentic commerce means
Mike separates three uses of AI in commerce: completing purchases, interpreting complex product searches, and improving retailers’ recommendations.
02:48 Why agentic commerce is accelerating
He describes the recent launches and retailer partnerships that have made the idea feel more immediate, alongside the access limits agents still face.
05:48 Buying versus shopping
Mike distinguishes a faster checkout from helping someone discover what they want. He sees particular value in handling a fuzzy brief and surfacing unexpected products.
08:38 Why Amazon restricts AI agents
His explanation includes returns, logistics, customer trust, and advertising revenue. He considers a possible future in which agents browse while people confirm the purchase.
12:08 Marketing to people and AI agents
The starting point is clearer product information and better brand governance. Mike also questions how much bot traffic represents real buying intent.
17:09 Why a strong brand still matters
Customers can ask for a brand by name. Mike also explores why agents might use familiar brands as shortcuts when making decisions, presenting this as his view of how the technology could develop.
22:54 Why brand consistency matters
Mike expects conflicting product claims to become a bigger liability. He argues for a coherent story across product pages, owned content, PR, and external coverage.
24:32 The future of performance marketing
Cheaper creative experimentation could give early adopters an advantage. Mike expects that advantage to narrow as more marketers use the same tools.
31:17 Early lessons from ChatGPT ads
Mike discusses a small campaign and conversations with other advertisers. He cautions against drawing broad conclusions from his test and sees room for better ad formats.
33:55 What advertising to agents could look like
He speculates that early formats could resemble offers, rewards, or affiliate marketplaces that agents consult when choosing a product or retailer.
36:03 Rethinking AI ad formats
The discussion considers interfaces beyond a static text ad, including moments when users are waiting for an AI task to finish.
39:24 Will Anthropic eventually run ads?
Mike offers a deliberately provocative prediction that Anthropic could eventually run advertising in a consumer product. He presents it as a view that few people share.
40:51 Will AI take over media buying?
He expects more execution to move into agents, while leaving open what new responsibilities media buyers might take on and where people will want control.
44:21 Why people still enjoy shopping
Travel browsing illustrates the pleasure of discovery and anticipation. Mike argues that useful agents should help people find new options while respecting the parts of choosing they enjoy.
47:15 What to expect at BRXND
Mike previews the November 5 gathering in New York and its central question: what happens to marketing when AI increasingly sits between brands and consumers?
Continue the conversation at BRXND NYC on November 5, 2026, at The Times Center. Event details and tickets.
