Marketing Trends

The episode contains genuinely novel frameworks - particularly the taxonomy of five machine customer types (delegated agents, autonomous buyers, co-buyers, multi-agent networks, intermediary brokers) and the application of BJ Fogg's behavior model to AI agents. However, substantial portions are devoted to conference recaps and exploratory thinking rather than distilled, actionable insights. The core insight density is strong but diluted by repetition and meandering tangents.

“One machine customer type. Yeah, like there's many types of machine customers and an AI agent is absolutely one of them. But it's not really big enough to describe all of the different types of machine customers that we, we are seeing. So I've identified about five.”

“Marketing to robots is not a thing. It doesn't have anything that we use our original marketing suite of tools to interact with the emotional narrative, the storytelling. It's like trying to market to a calculator.”

Katya Forbes presents genuinely fresh angles: the explicit rejection of 'marketing to robots' as a category, the machine customer taxonomy itself, and the insight that emotional inference in cars paired with agentic commerce creates new vulnerability surfaces. The China field research and specific examples (Alibaba's 120M bubble tea orders, Bank of Bots, Mercedes Pay) provide grounding. However, some frames (dual-track marketing, values-driven differentiation) have circulated in adjacent conversations.

“Alibaba ran a campaign to give people 25 yuan coupons in order for them to use the Chen AI model that Alibaba has created to customize and order bubble tea...they did 10 million orders in the first nine hours and 120 million orders across the life of that campaign.”

“trust is a differentiator in the marketplace, particularly as younger generations are buying in alignment with their values and statements about their values are becoming part of the descriptor that they're giving to their AI agents”

Katya Forbes is the founder of a consulting practice and author of a field guide on machine customers, giving her relevant domain expertise and original research. However, the transcript reveals her primary evidence is self-directed travel observations and academic paper summaries rather than operational execution at scale. She is more of a thought-leader/researcher than a proven operator who has built or scaled a revenue-generating business in the space she's advising on.

“I am very easy to find, so I work out loud on this. I work out loud on LinkedIn, I'm easy to find there as well as on Substack, uh, where I write as the CX Evolutionist. My book obviously is available for people”

“I visited Alibaba, Tencent Unitree Robotics and byd...I had an opportunity to interrogate their VP of marketing and one of their product people”

Strong specific references: Alibaba's 120M orders, 25 yuan coupons, spring festival timing, specific robotics vendors (Unitree, Boston Dynamics), BYD's intelligent cockpit features, Mercedes Pay (2023), Worldpay survey (8k respondents, 65% China vs. 53% France), Walmart-Pactum deal. However, many claims lack hard numbers: emotional inference capabilities are described but not quantified; the 'five machine customer types' are named but lack concrete adoption metrics; Bank of Bots' loan is mentioned without deal size.

“they did 10 million orders in the first nine hours and 120 million orders across the life of that campaign”

“Worldpay did a really great survey just at the end of 2025, ah, with about 8,000 people across all sorts of different countries. Um, and in China, 65% of people said I'm already shopping with an AI agent”

The host asks strong opening questions ('what did you see that shifted you?') and shows genuine curiosity about specifics (Alibaba campaigns, Mercedes tech, consumer trust dynamics). However, follow-ups often veer into softballs or abstract agreement rather than productive pushback. When Katya makes speculative claims ('I can see a possible future'), the host doesn't press on evidence or likelihood. The host also allows extended monologues without interrupting for clarity, and rarely challenges assumptions.

“Katya, welcome to Marketing Trends...I'm excited because you are such a unique guest and even on our prep call when I was talking to you, I'm like, oh, man, so many good, juicy ideas and contrarian takes.”

“So registered though, I'm like, I've made agents, I don't know what it means to be registered. So is this just for like enterprise companies who know how to go through that hoopla to make it registered?”