Every missed call is a lost customer: why STL restaurants need voice AI
The phone rings during the Friday rush. Nobody answers. The customer calls someone else. Here's what OpenTable and Resy don't solve, and what voice AI does.
A St. Louis restaurant we work with ran a 72-hour call log. Total inbound calls: 312. Total answered: 141. Total that resulted in a reservation, takeout order, or catering lead: 89. The rest hit voicemail or rang out.
The missed calls didn't call back. They called the next restaurant on their list. At an average ticket of $52, that's roughly $8,500 in revenue that walked to competitors in three days. Multiply across a year and it's six figures.
The phone ringing and nobody answering is a restaurant-specific problem because the rushes are predictable. The lunch push from 11:30 to 1. Friday night 5 to 8. Valentine's Day, Mother's Day, the week of prom. Your host can't keep up, and the calls that should have been reservations become walkouts.
OpenTable and Resy are not solving this. They solve the reservation-by-app use case, which is real but partial. The caller asking whether you take a party of 12 at 7:45 on Saturday isn't using the app. They're calling. If nobody picks up, they're gone.
What voice AI does in a restaurant context. It answers every call. It checks the POS and availability. It holds a reservation, adds the party size and dietary notes, and texts a confirmation. It takes a takeout order and writes it directly to the kitchen ticket system. It handles the 'are you open,' 'where are you,' 'do you have a gluten-free menu' calls that eat your host's attention.
Integration reality. A good voice agent for restaurants integrates directly with OpenTable, Resy, Toast, Square, or whichever stack you run. Not a Zapier bridge. A direct API. The reservation shows up on the floor plan in real time. The takeout order prints at the line.
The specific numbers for an STL restaurant: at 300 inbound calls a week with a 50% miss rate, you're losing roughly 150 opportunities. A voice agent captures 80-90% of those. At a 30% conversion to actual revenue and an average ticket of $50, that's $1,800-$2,100 a week in recaptured revenue. Monthly run rate on the agent: $400. Payback: week one.
The guest experience is the part restaurateurs worry about. They shouldn't. A properly tuned voice agent is faster and more accurate than a harried host on a Friday night, and it hands off to a human if the caller asks for one. The guests that would have cared are the minority. The guests that would have hung up on a phone-tree robot aren't dealing with a phone tree.
Restaurants in STL already running this: a dozen. They're not talking about it publicly because it's a competitive advantage and they know it.
What to do next if you run a restaurant and your phone rings during service: pull a week of call log data. Count the rings that hit voicemail. Multiply by your average ticket. If the number is over $2,000 a week, a voice agent pays for itself faster than you finish reading this article.