What's Happening?
DoorDash CEO Tony Xu announced during the company's third-quarter earnings call that third-party AI agents have generated only several hundred orders in total since their inception. This figure stands in contrast to the 993 million orders processed by
DoorDash last quarter, indicating a very low adoption rate for AI-initiated transactions on the platform. Xu noted that consumers predominantly continue to use DoorDash's own app for routine orders, although the company is exploring collaborations with external AI providers. This disclosure aligns with similar reports from other delivery services, suggesting that the ability to place orders through external AI assistants has not yet resulted in substantial transaction volumes for these platforms. The company's stance is to not 'fight gravity,' implying a pragmatic approach to integrating new technologies based on user behavior.
Why It's Important?
The limited transaction volume from third-party AI agents on platforms like DoorDash highlights a significant challenge for the broader e-commerce and AI industries. Despite the growing interest in AI-driven shopping tools, consumers appear to prefer direct interaction with established apps for ordering and delivery services. This trend suggests that for AI agents to gain widespread adoption, they must offer a demonstrably superior or more convenient experience than existing, already efficient platforms. For businesses investing in AI, this indicates a need to focus on enhancing user experience and providing clear value propositions rather than simply offering AI as a novelty. The slow uptake could impact investment strategies in AI development for transactional purposes, prompting a re-evaluation of how AI can genuinely improve the customer journey in the delivery sector.
What's Next?
DoorDash, despite the low current adoption, plans to continue exploring ways to work with external AI providers, suggesting a long-term view on the potential of AI. The company's strategy will likely involve identifying specific use cases where AI agents can offer unique value or streamline processes that are not currently optimized by their native app. For the broader industry, this situation may lead to a shift in focus from agent-initiated transactions to AI-assisted shopping, where AI helps with product discovery or cart building, but the final purchase decision and execution remain with the user through the platform's native interface. Future developments will likely involve more sophisticated AI integrations that seamlessly blend into existing user workflows, rather than attempting to replace them entirely, to overcome current adoption hurdles.
Beyond the Headlines
The low adoption of third-party AI agents for direct transactions on platforms like DoorDash raises questions about consumer trust, data privacy, and the perceived autonomy in purchasing decisions. While AI can offer convenience, users might be hesitant to delegate full purchasing power to an external agent, preferring to maintain control over their orders, especially concerning details like substitutions or special instructions. This could also reflect a broader societal comfort level with AI, where users are more accepting of AI as an assistant rather than a fully autonomous agent. The ethical implications of AI agents making purchasing decisions on behalf of consumers, including potential biases or errors, could also be a contributing factor to the slow adoption, prompting a need for greater transparency and user control in AI-driven commerce.
