Restaurant Missed Call Text-Back in Louisiana

Missed calls fire a webhook into n8n. Opening SMS within sixty seconds. AI agent runs the reservation conversation: date, time preference, party size, contact, special occasion or seating preference. Booking writes to the reservation platform. Modifications and cancellations handled by phone-number lookup. Common questions (hours, menu, dress code, parking) get pre-approved answers from the knowledge base.

A real exchange looks like this. It is 6:47pm on a Thursday at Bistro Lumiere. The host is double-seating the Thursday night rush. Jessica calls because she wants to book a Saturday 7:30pm dinner for four to celebrate her father's birthday. The call rings out. At 6:48pm Jessica gets an SMS: 'hi, this is the host stand at Bistro Lumiere, sorry we just missed your call. Were you looking to make a reservation?' Jessica replies 'yes, Saturday 7:30pm for four, my dad's birthday.' The agent checks Saturday availability via the OpenTable API and sees 7:30pm is booked but 7:45pm and 8pm are available. It offers both options, captures the birthday note for the kitchen, asks if Jessica has any seating preference (window if available), and asks about dietary restrictions (one guest is gluten-free). Jessica picks 7:45pm. By 6:51pm the reservation is on the books, Jessica has a confirmation with the time and a note that the kitchen will be informed about the birthday and the gluten-free guest, and the restaurant captured a four-top that would have moved to a competitor inside the hour.

The AI's qualification flow is restaurant-specific in ways that matter for both reservation accuracy and guest experience. It distinguishes between standard reservations (single tables under the configured size threshold), large-party inquiries that require manager involvement, and event or buyout inquiries that need to route to the private dining manager. It captures special occasions (birthdays, anniversaries, business dinners, first dates) because the kitchen and front-of-house team use this information to deliver appropriate hospitality moments. It captures dietary restrictions and allergies, writing them into the reservation note so the kitchen can prep accordingly. It handles modifications and cancellations through phone-number lookup against the reservation system, so existing reservations can be moved or canceled without the guest having to call back. It also handles informational queries (operating hours, menu questions, dress code, parking, accessibility, whether dogs are allowed on the patio) through a knowledge base that the manager configures during setup, which deflects roughly half of the inbound phone volume away from the host stand entirely.

The host stand is staffed for cover count, not for call volume. Friday night dinner service has the host doing maybe one hundred guest interactions with the phone ringing thirty times. The misses are real, the reservation revenue lost is real, and most restaurants do not have a fix because they cannot afford a dedicated reservation desk. The agent provides it.

The structural labor problem in restaurants is the host-stand workload during peak service. A typical mid-sized restaurant runs with one host during lunch and one to two hosts during dinner, plus a manager who flexes in to help during peak rushes. The host job during dinner service includes greeting guests, managing the wait list, seating tables, clearing entries, coordinating with servers about table turns, fielding to-go orders, handling guest complaints, processing pickup checks for takeout, and answering the phone, all simultaneously. The phone is the lowest-priority task because the guest at the door is right there making eye contact and the guests at the wait-list station are watching for their table. So the phone rings out, the reservation goes to a competitor, and the host never knows the lead existed because there is no time to listen to voicemails until after service. Hiring a dedicated reservation desk staffer costs forty to fifty thousand annually plus benefits, and most restaurants cannot justify it because reservations also flow through the website and OpenTable.

The second structural piece is the timing pattern of reservation inquiries. Most reservation calls for a weekend service come in during the week, with Wednesday and Thursday between 11am and 8pm being the peak. That overlaps directly with lunch service prep, lunch service itself, and dinner service prep, all of which fully occupy the host stand. The result is that the most economically important reservation calls of the week hit voicemail at the worst possible time, and the host has no realistic way to call back before the diner has booked elsewhere. The restaurants that figure out fast text-back response capture this weekday-call-for-weekend-service pattern and grow their reservation revenue significantly. The restaurants that do not figure it out leak the highest-value reservations of the week to whichever competitor answered first, and they cannot easily diagnose the lost revenue because the missed calls do not appear in any sales report.

Average check at casual is forty to sixty per guest. Fine dining is a hundred to two-fifty. A four-top captured is one-sixty to a thousand in revenue with strong contribution margins. A restaurant capturing ten extra reservations a week is sixteen thousand a month in incremental revenue.

The expected-value math breaks down by reservation type in ways that make the recovery economics specific. A standard two-top reservation at casual pricing generates eighty to one twenty in food-and-beverage revenue and accounts for about thirty-five percent of inquiries. A four-top accounts for forty percent and generates one-sixty to two-forty at casual or four to six hundred at fine dining. Larger parties (six-plus) account for fifteen percent and generate three hundred to over a thousand in revenue, with higher conversion to additional revenue lines like specialty wine selections, dessert courses, and group celebration add-ons. Special-occasion reservations (anniversaries, birthdays, business entertaining) account for the remaining ten percent and skew higher in spend because the host has a reason to celebrate. Run those weights against twelve recovered reservations weekly with average four-top equivalents and the expected revenue lands between twenty-five hundred and ten thousand dollars per week added, depending on the restaurant's check average.

n8n workflow with Twilio. AI agent prompt for restaurant text conversations with reservation booking and FAQ handling. Reservation platform integration. Knowledge base configuration. Setup guide.

The integration options ship to cover the dominant reservation platform landscape. The missed-call trigger works with Twilio (default), CallRail, JustCall, and the call-tracking features built into RingCentral. The reservation booking integrates natively with OpenTable through their partner API, Resy through ResyOS, SevenRooms (which dominates higher-end restaurant groups), Tock (preferred by chefs and prix fixe concepts), and Yelp Reservations. For restaurants without a reservation platform, the workflow can write to a Google Calendar mirror that the host stand reviews each shift. The knowledge base for FAQs (hours, menu, dress code, parking, accessibility, dietary policies, group reservation policies) is configured per restaurant during setup and lives in a Notion or Airtable document that the manager can update without engineering help. Optional integration with the restaurant's POS for special-event ticketing (wine dinners, chef's table, holiday menus) so the agent can sell tickets directly through the text conversation.

The prompt is the deepest part of the template and has been refined against roughly two hundred deployed restaurant conversations across casual and fine-dining concepts. The system prompt includes explicit guardrails: never modify pricing or quote firm prices because pricing is set by the menu, never make food preparation promises that the kitchen has not committed to (no, the chef cannot prepare an off-menu dish without prior arrangement), never modify or cancel a reservation made through a third-party platform without confirming with the host, always handle dietary restrictions with appropriate seriousness because allergens can be life-threatening, always disclose if the restaurant cannot accommodate a specific request, never engage with reservation requests for events the restaurant does not offer (the agent gracefully redirects to private dining for group inquiries above the configured threshold). The prompt also handles edge cases that broke earlier versions: callers asking about menu items that are seasonal and may not be currently available, callers asking about wine list specifics, callers asking about gift card purchases, and callers asking about the chef or restaurant's affiliation.

Louisiana has 4.6 million residents distributed across major metros including New Orleans, Baton Rouge, Shreveport, Metairie, and Lafayette. Louisiana's hurricane vulnerability shapes its home services markets dramatically. Insurance market dynamics drive roofing and water mitigation demand. New Orleans has unique housing stock and pest control challenges.

The seasonality of restaurant work in Louisiana is the single biggest factor that shapes how this missed call text-back actually performs in the market. Hurricane season (June through November) creates massive demand spikes. Year-round warm climate. The template's qualification logic, dispatch rules, and conversation flow are tuned to handle these patterns rather than forcing the agency operator to customize from scratch. Shops that deploy this in Louisiana markets see the seasonality framing show up in the conversations from the first call.

Regulatory framework for restaurants in Louisiana varies at the local level rather than statewide, which is worth understanding because licensing references in customer conversations need to match local jurisdiction. The agent template handles this correctly by deferring licensing-specific questions to local context rather than asserting state-level rules that may not apply.

A day. OpenTable and Resy have partner APIs. The most important customization is the restaurant's voice. Forty-five minutes with the manager. Test against a personal phone. Agency operators charge eight hundred to fifteen hundred for setup and four hundred to six hundred a month.

The ongoing tuning in restaurants follows a quarterly rhythm aligned with menu changes and seasonal service patterns. For the first ninety days, pull conversation logs weekly and review for two patterns. First, reservations that booked but produced no-shows or modifications that suggest the agent failed to set proper expectations. Second, conversations that dropped off before booking because the agent could not handle a specific question. Common findings include the agent under-promoting prix fixe or tasting menu options that the kitchen wants to highlight, the agent missing dietary restrictions when guests use casual language ('I don't really eat dairy'), and the agent failing to recognize VIP-customer signals when regulars call from new phone numbers. Adjust monthly during the first quarter, then settle into quarterly tuning aligned with menu changes, holiday service patterns, and any seasonal menu launches. After about six months the prompt is well-tuned for the specific restaurant's voice.