How a Málaga restaurant saved 15 hours a week with AI

Carlos has been running his restaurant in central Málaga for 12 years. Good food, solid team, loyal customers. But when he called us in October 2024, he had a problem keeping him up at night: too much of his team's time spent managing bookings by phone and WhatsApp.

"I worked out that between me and the floor manager, we spent about 3 hours a day just answering booking messages. Weekends were chaos — the phone wouldn't stop while we were trying to run the service."

The diagnosis: what was really happening

When we analyzed the situation, we found three concrete problems:

• Bookings lost after hours. Most messages came in between 10 PM and 11:30 PM — when the restaurant was already closed and nobody was answering. Many of those customers didn't try again.
• Repetitive questions. 70% of the messages were variations of the same five questions: hours, availability, menu price, allergy info, parking.
• Cancellations with no rebooking. When someone cancelled a booking, the table stayed empty because nobody was actively managing the waitlist.

The solution: AI Agent on WhatsApp

We implemented a Bissar AI Agent connected to the restaurant's WhatsApp Business number and to the booking system they were already using. The agent was configured with:

• All the restaurant's information: hours, menus, prices, booking policy, allergy information
• Real-time access to the booking calendar to check availability
• Ability to confirm, modify, and cancel bookings directly
• Active waitlist: when someone cancelled, the agent automatically offered the slot to the first people on the list
• Automatic reminders 24 hours before each booking

How the implementation went

A real conversation at 11:14 PM

That conversation would have gone unanswered until the next morning — and the customer would probably have booked somewhere else. With the agent, it was confirmed in 90 seconds at 11 PM.

The results after one month

A month after implementation, we measured the results. The numbers were clear:

• 15 hours saved per week between Carlos and the floor team — the equivalent of nearly two workdays a month.
• +28% confirmed bookings compared to the previous month, thanks to night coverage and the active waitlist system.
• No-show rate down to 4% (previously 12%) thanks to the automatic reminders.
• €9,000 estimated annual savings from time freed up plus the increase in bookings served.

"The biggest change wasn't the hours saved — it was no longer being glued to my phone during service. Now I can focus on what matters: the customers right in front of me."

— Carlos M., owner of Restaurante La Bodega · Málaga Centro

What we learned from this case

Three things we saw clearly in this project and applied to every one after it:

Tone matters as much as functionality. The first agent we set up sounded too corporate for a neighborhood restaurant. Once we tuned the tone to something warmer and more local, conversation quality improved by 15%.

Always define when it should escalate to a human. The agent isn't meant for everything. Carlos decided the agent would always pass direct complaints, bookings for groups of more than 10, and special event menu requests to a human. That avoided awkward situations.

The active waitlist is gold. That module Carlos initially thought he wouldn't use turned out to be one of the most valuable. On weekends, the agent handled an average of 3-4 cancellations and rebooked them automatically, recovering revenue that used to just be lost.