Basically no one is using AI agents for their DoorDash orders

But it's still early days for the relatively new e-commerce use case.

The top 3

Discrepancy Between Internal AI Tools and Third-Party Agents: DoorDash’s internal AI tools, such as 'Ask DoorDash', have proven successful by delivering 50% higher basket values and 60% more unique items compared to traditional manual orders. In contrast, third-party agents struggle because food ordering is a high-frequency, low-consideration task that consumers prefer to execute directly via established, trusted interfaces.
Early Infrastructure and Market Constraints: The low volume is partially attributed to the nascent state of integration; DoorDash released its Model Context Protocol (MCP) connector less than a month ago, and it is currently limited to a small set of beta partners. Unlike 'long consideration' purchases where agents could proactively provide value, simple food ordering lacks the complexity where an external agent's overhead provides a clear competitive advantage over native apps.