OhMiam: a real-world restaurant digitalisation case study by Hebora.
The real problem in a restaurant in 2026
Restaurant owners don't lose customers because of the food. They lose them on the surrounding details: the reservation we couldn't take because we were mid-service, the Google review that drops because no one replied, the Saturday night no-show that destroys margin.
The trap is thinking more staff fixes this. The real point is to separate what needs a present human (welcoming guests, suggesting dishes, reading the table) from what doesn't (confirming a reservation, the day-before reminder, replying to 'are you open tomorrow?'). Everything else is a fake technical debate.
Useful automation in a restaurant never adds a screen to monitor during service. It disappears. If you can see it, it's badly built.
What is worth automating in a restaurant
Online reservations with instant confirmation (Zenchef, Bookatable, or a custom module if you do more than 50 covers/day). Day-before SMS reminders to break no-shows. An automated FAQ for hours, menu, allergens — that's 70% of your daily phone calls.
QR code takeaway ordering: essential since 2021 — almost every restaurant that survived COVID adopted it. Google review follow-up with a discreet nudge after the visit. A monthly newsletter focused on loyal customers, never on walk-ins.
Every automation candidate must pass a simple test: does it free time in the kitchen or on the floor during service? If the answer is 'it adds noise', you're on the wrong tool.
Step-by-step method to automate a restaurant
<strong>Step 1.</strong> Map for one week when the phone rings, who calls about what, and how long it takes. You'll find that 60% of calls cover 5 identical questions.
<strong>Step 2.</strong> Get reservations online. Not a contact page — a real module with slots, durations, group sizes. Test for 2 weeks with a human fallback.
<strong>Step 3.</strong> Activate D-1 SMS reminders on reservations. The biggest no-show lever, especially for parties of 6 or more.
<strong>Step 4.</strong> Write a structured FAQ (hours, allergens, terrace, parking, payment). Wire it to a simple AI assistant that filters routine questions and hands over for the rest.
<strong>Step 5.</strong> Set up the QR code menu and takeaway order flow, with a clean kitchen-side workflow (direct ticket print).
<strong>Step 6.</strong> Measure. How many fewer calls? How many fewer no-shows? How many additional takeaway covers? If ROI isn't there at week 8, the tool or the configuration is the issue.
The five automations that pay off the most
<strong>1. D-1 SMS reminders on reservations.</strong> Cost: 30-50 €/month based on volume. Typical gain: -40 to -60% on no-shows.
<strong>2. Online reservation module with auto-confirmation.</strong> Cost: 0 (Google Reservations) to 80 €/month (Zenchef). Gain: 2-5h/week of phone-desk time saved.
<strong>3. Automated FAQ wired to WhatsApp Business or the site widget.</strong> Cost: 30-100 €/month depending on solution. Gain: -50% on pure-question phone calls.
<strong>4. Google review follow-up after the visit.</strong> Cost: 20-40 €/month (Trustmary, Pluspoint). Gain: 2-3x on the volume of reviews collected.
<strong>5. Click & collect or QR takeaway ordering.</strong> Cost: 50-150 €/month or per-order commission. Gain: a new revenue channel, especially weekday lunch.
What to avoid in a restaurant
No badly scoped chatbot that answers nonsense in your name. A chatbot that invents a menu or hours drives customers away faster than it attracts them. If you don't have time to brief it well, don't deploy it.
No automation that overrides hospitality. No one goes to a restaurant to talk to a robot. AI can take reservations and FAQ — the moment of arrival and the floor service stay human, full stop.
No over-collection of customer data without value. Asking for date of birth, phone, allergies and food preferences on a Sunday lunch reservation form is no. GDPR requires you to justify purpose and limit retention. Be strict.
Realistic costs and ROI to automate a restaurant
For a neighbourhood restaurant (30-60 covers/day), expect 80-250 €/month combined tool spend (reservations + SMS + FAQ + reviews), plus an initial scoping fee of 800-2 500 € so you don't end up with 4 tools that talk badly to each other.
Typical ROI shows in saved phone-desk time (5-10h/week), no-shows avoided (1-3 covers/week recovered at 35-50 € average), and additional Google reviews. Concretely, a properly automated restaurant pays back the scoping in 3-5 months and frees a half-time front-desk role for something more useful.
FAQ
Should a small restaurant automate reservations?
Yes in most cases. As soon as you handle more than 20 covers/day, the cost of manual handling (phone, paper diary, errors) exceeds a 30-80 €/month tool. Below that, a free module like Google Reservations is enough. Above 80 covers/day, a dedicated tool (Zenchef, The Fork) becomes essential.
How do you cut no-shows in restaurants?
Three levers. D-1 SMS reminders (most effective, -40 to -60% no-shows observed). Symbolic 5-10 €/person deposit for parties of 6+. Clear cancellation policy shown at booking. Combining the three typically cuts no-shows by 3-4x.
What does it cost to automate a restaurant in Brussels?
Indicative range: 80-250 €/month in tools depending on size and features. Hebora scoping fee between 800 and 2 500 € depending on scope (one simple tool or full orchestration). For most Brussels restaurants, the total first-year investment is between 2 500 and 5 500 €.
Can a chatbot replace a restaurant's front desk?
No, and no serious provider should promise that. A well-scoped chatbot filters simple questions (hours, menu, allergens, terrace) and hands over to a human for the rest. It doesn't take a complex reservation, doesn't handle a complaint, doesn't suggest a dish. It frees your front desk, it doesn't replace it.
Are these automations GDPR-compliant?
Yes when properly scoped. EU hosting, signed DPAs with vendors, limited retention (12 to 24 months on customer contact data), clear legal basis (consent for the newsletter, contract execution for reservations). Hebora frames this before going live, not as a patch later.
What's the difference between Zenchef, The Fork and a custom tool?
Zenchef and The Fork are off-the-shelf platforms — easy to plug in, but they take a commission or subscription and expose your customer data to their ecosystem. A custom tool (what Hebora builds for high-volume restaurants) costs more upfront but keeps the data with you, integrates with your POS, and costs less to run beyond 80-100 covers/day.
How do you automate QR-code takeaway?
Three steps. (1) Map the takeaway menu (often a reduced version of the dine-in menu). (2) Pick a simple module (OhMiam, Sunday, or custom if you want full control). (3) Wire to the kitchen via a dedicated printer so the floor service stays smooth. End-to-end deployable in 2-4 weeks.
Does a restaurant need a newsletter?
For a destination restaurant (fine dining, strong concept, event-driven brunch), yes — once a month is enough, focused on novelty and events. For a walk-in neighbourhood spot, marginal. Better invest the time in Google Business Profile and the My Business listing than in a newsletter that ends up in spam.
Want to automate your business?
Hebora helps small and medium businesses in Brussels scope their automations before touching any tool. 30 free minutes to figure out what's actually worth automating in your context.
Talk to Hebora