The DoorDash Problem: How AI browsers are a huge threat to Amazon
Timestamps are as accurate as they can be but may be slightly off. We encourage you to listen to the full context.
Podcast Summary
In this episode of Decoder, Eli Patel explores what he calls the "DoorDash problem" - the fundamental challenge of what happens when AI agents get between service providers and customers. (00:35) The episode centers around Amazon's recent lawsuit against Perplexity to prevent its AI-powered Comet browser from shopping on amazon.com, marking the first major battle in the AI browser wars. (02:17) Patel examines how this shift from smartphone apps to AI agents could transform the internet economy, potentially reducing companies like DoorDash, Uber, and Amazon to mere commodity databases competing solely on price while losing valuable customer relationships and monetization opportunities.
- Main Theme: The economic disruption potential of AI agents acting as intermediaries between consumers and service providers, threatening traditional app-based business models that rely on direct customer relationships for advertising revenue, subscriptions, and cross-selling opportunities.
Speakers
Eli Patel
Eli Patel is the editor-in-chief of The Verge and host of the Decoder podcast. He focuses on big ideas and complex problems in technology, with particular expertise in analyzing how emerging technologies impact business models and the broader economy. Patel has been tracking the intersection of AI and commerce for over a year, conducting interviews with numerous tech CEOs about the potential disruption of AI agents.
Key Takeaways
AI Agents Threaten Traditional App Monetization Models
The core issue isn't just about convenience - it's about economics. (00:57) When AI agents interact directly with services like DoorDash or Amazon, they bypass all the ways these companies traditionally make money beyond the core transaction. User reviews, advertisements, loyalty programs, upsells, and partnerships become irrelevant because AI agents don't engage with these elements. This could transform profitable service companies into commodity providers competing solely on price, potentially making their business models unsustainable even if technically superior.
Amazon Has the Most to Lose in an AI-First World
Amazon's aggressive legal action against Perplexity makes strategic sense when you consider their vulnerabilities. (24:41) The company generated $17.7 billion in advertising revenue in their most recent quarter - a 24% year-over-year increase. AI agents that shop for users won't see these ads or care about sponsored placements. Additionally, Amazon has already commoditized many of its products through third-party sellers, making it easier for AI agents to find equivalent items elsewhere based purely on price, potentially undermining both their ad revenue and Prime subscription value.
Build Experience First, Economics Second
Uber CEO Dara Khosrowshahi provided a pragmatic approach to dealing with AI agents. (16:37) Rather than trying to figure out pricing models upfront, he advocates for charging zero initially to test whether AI partnerships provide incremental value or simply cannibalize existing customers. This allows companies to optimize the user experience first, measure real impact, and then negotiate fair economics based on actual value delivered rather than theoretical concerns.
Brand Differentiation Provides Natural Protection
Several service company CEOs expressed confidence that their deep operational moats protect them from AI disruption. (12:30) ZocDoc's CEO Oliver Kurtzig emphasized their 20+ years of healthcare system integration, while TaskRabbit's CEO highlighted their vetted network of background-checked workers. The argument is that AI agents still need access to real-world infrastructure and trusted service networks - things that can't be easily replicated by AI companies trying to build competing services from scratch.
The Legal Framework for AI Agents is Unclear
Perplexity's defense raises fundamental questions about AI agent permissions and digital rights. (28:47) The company argues that "user agents are exactly that, agents for the user" and should have the same permissions as the human they represent. This "software as labor" perspective challenges traditional terms of service and could reshape how we think about automated interactions with websites. However, Amazon's lawsuit under the Computer Fraud and Abuse Act suggests this legal territory is far from settled.
Statistics & Facts
- Amazon reported $17.7 billion in advertising revenue in its most recent quarter, representing a 24% increase year-over-year. (24:41) This puts Amazon on track to generate more than $60 billion in ad revenue in 2025 alone, making it the third-largest ad business among tech giants behind only Google and Meta.
- Amazon's advertising business has grown so substantially that it's now positioned to rival AWS as a major profit center for the company. (25:08) The company is not only using AI to help brands generate images, product descriptions, and video ads, but is also selling the technology that powers its ad platform to other retailers.
- Perplexity launched its AI-powered search engine in December 2022, shortly after ChatGPT's debut. (27:18) Since then, the company has faced multiple lawsuits from content providers including Dow Jones (Wall Street Journal), Encyclopedia Britannica, Merriam-Webster, and Reddit for scraping and using their content without permission.
