What changed in 2026
- Square added restaurant ordering to ChatGPT and Claude in July; eligible U.S. restaurants on Square Online Ordering were opted in, and orders land in the same POS and kitchen display as other online orders.
- Google brought ordering into Ask Maps in Google Maps in August, starting with Square and Toast restaurants.
- Papa Johns became the first partner for Google Cloud's food ordering agent.
- Starbucks launched a ChatGPT app in beta that hands guests to its own app to check out.
- Retail is moving too: Walmart announced Agent Profiles, an Agent Gateway and terms for AI agents, so customers can shop through the assistants they choose. The same six areas below apply to any catalog.
The six areas I assess
- Menu as data: every item, size, modifier, required choice, price and store-level availability, written down rather than enforced by screens.
- Where checkout happens: inside the agent, or handed back to your app, and what each costs you.
- Loyalty recognition: members identified without a password, earning points and reordering "the usual" in any channel.
- Offer rules: eligibility and stacking a machine can apply correctly.
- Wrong-order safeguards: read-backs, required choices, and support that can see what the agent was told.
- Channel reporting: agent orders measured separately from app and web.
Why me
I've built the systems these agents depend on: Sonic's Digital Menu Manager, which powered Order Ahead across every Sonic location, and the menu, offer and identity work behind the Jack in the Box app. I've also built an MCP server for fast-food ordering, so I know where agents break.
How we'd work together
As a Digital Audit focused on agent readiness (2–3 weeks, starting at $8,500), ending in a prioritized list of what to fix first. I can also build the pieces, like a menu data feed, an agent-ready ordering pilot or an MCP server, as Design & Build, scoped per project.