Casey's Voice AI Rollout: Lessons for Phone Orders

Casey's General Stores has expanded its partnership with SoundHound AI to more than 2,600 locations, using AI voice agents to take pizza and food orders by phone (as announced by SoundHound AI, April 23, 2026). The agents are trained on Casey's own menu and have already handled more than 21 million guest interactions. For a small restaurant or convenience store, the news matters less for its scale than for what it proves: phone orders taken by AI, at high volume, without losing the order.

This post covers what Casey's actually deployed, how the technology works during a call, and what a much smaller restaurant or store can take from it without needing a national chain's budget.

Key Takeaways

• Casey's, the United States' third-largest convenience retailer and fifth-largest pizza chain, expanded its SoundHound AI voice ordering agents to more than 2,600 stores (as announced by SoundHound AI, April 23, 2026).
• The agents have handled more than 21 million guest interactions to date and are trained specifically on Casey's menu so orders come back accurate.
• The stated goal is fewer missed orders and shorter hold times during peak meal periods, when a single cashier cannot answer the phone and run the register at once.
• Voksha's own analysis of call center benchmarks found that businesses answer only 37.8% of inbound calls on average (411 Locals, 2024, After-Hours Economy report), which is the same failure mode Casey's is trying to close with its rollout.
• A single-location restaurant does not need enterprise voice infrastructure to solve the same problem; a phone-answering AI receptionist can cover the same peak-hour gap for a fraction of the cost.

What Casey's Actually Rolled Out

SoundHound AI's voice ordering agents now take calls at more than 2,600 Casey's locations, up from a smaller footprint under the companies' earlier partnership (as announced by SoundHound AI, April 23, 2026). The agents are engineered to understand natural speech, so a caller can order "a large pepperoni, extra cheese, and a two-liter Pepsi" the same way they would to a person behind the counter, rather than navigating a phone tree. Sanjeev Satturu, Casey's SVP and Chief Information Officer, said the goal is to let store teams work more efficiently across thousands of locations while keeping order quality consistent (as announced by SoundHound AI, April 23, 2026).

Beyond taking orders, the agents can answer menu questions and describe current promotions, which is the part of a phone call that most often causes a human cashier to put a customer on hold. During busy meal windows, that hold is exactly when convenience stores and small restaurants lose orders to a caller who simply hangs up.

Why Peak-Hour Answering Is the Real Problem

The story behind the Casey's expansion is not really about AI sophistication. It is about the gap between how many calls come in and how many a store can physically answer while also serving the counter, the drive-through, and delivery pickups. Voksha's After-Hours Economy research, drawing on third-party call center data, found that businesses answer only 37.8% of their inbound calls on average, and that 80 to 85% of callers who reach voicemail simply hang up rather than leave a message (After-Hours Economy report, citing 411 Locals, 2024). A restaurant with one phone line and a busy kitchen faces the exact version of this problem that Casey's is solving at scale.

What Small Restaurants Can Learn Without Casey's Budget

A single-location restaurant or a small convenience store cannot build custom voice AI trained on years of order data. But the underlying fix, an AI receptionist that answers every incoming call, takes the order or the reservation request, and hands off anything it cannot resolve to a person, is available off the shelf. Voksha's guide to restaurants using AI to handle reservations and takeout by phone walks through how a smaller operation sets this up without hiring a second phone-answering employee for the dinner rush.

The lesson from Casey's scale is proportional, not literal: a restaurant does not need 2,600 locations of order data to benefit from an AI agent that never lets a phone ring unanswered during the rush.

What This Means for Small Businesses

If your business takes phone orders or reservations, the busiest hour of your day is also the hour you are least equipped to answer every call. Casey's data point, more than 21 million interactions handled without adding headcount, is a signal that AI voice agents have matured past novelty for order-taking specifically. A small restaurant, pizzeria, or convenience store considering the same move does not need to build anything: an AI receptionist configured with your menu and hours can answer calls the moment your line is busy, quote wait times, and take basic orders, escalating anything unusual to your staff. That is the same problem Casey's solved, just sized for one location instead of 2,600.

Sources

• Casey's and SoundHound AI announcement, April 23, 2026
• Yahoo Finance coverage of the Casey's-SoundHound expansion, April 23, 2026
• Voksha After-Hours Economy report, citing 411 Locals (2024)

Published by the Voksha team. Voksha builds AI receptionists for small businesses. No Voksha call data specific to this event is used in this article; the SoundHound figures are as announced by the company.

Frequently Asked Questions

Does an AI receptionist only work for restaurants with a big IT budget?

No. Off-the-shelf AI receptionists are built to be set up by an owner or manager in minutes, without custom development, unlike an enterprise deployment like Casey's.

Can a small restaurant's AI agent take full orders like SoundHound's?

Order-taking depth varies by vendor. Most small business AI receptionists focus on answering, quoting wait times, taking reservations, and capturing takeout requests, then confirming details with a human if the order is complex.

Ready to stop losing calls during your own rush hour? See Voksha's pricing and get an AI receptionist live in minutes.