Reservation Agent: bookings with no no-shows and no manual calls
HoReCa · 12 restaurants
n8n · iiko
Rollout in 26 days
Payback in 2.1 months
Booking
n8n
Messenger
Confirmed
AI call
Waitlist
The brief
What the client came with
The chain was losing revenue to no-shows while the host spent hours calling guests.
PROBLEM 01
No-shows = 22% of bookings
Almost a quarter of guests neither showed up nor cancelled. The table sat idle and the revenue was gone.
PROBLEM 02
≈$5.4K a month in lost revenue
Across a chain of 12 venues, no-shows and late cancellations ate about $5.4K a month.
PROBLEM 03
3 hours a day of the host calling guests
Confirmations and reminders were made by phone by hand, and only about half the guests answered.
Solution
What we automated
We connected iiko to the messaging channels so the pipeline runs bookings, not the host.
n8n + iiko
Booking confirmation
Multi-channel: Telegram, then WhatsApp, then SMS. The guest confirms themselves and the status flows into iiko.
AI call
Voice call
If a guest does not reply in the messenger, the assistant calls and confirms the booking by voice.
Waitlist
A freed table
A cancellation or a no-show hands the table straight to the waitlist, and the dashboard shows the occupancy forecast.
Booking
iiko
n8n
orchestration
Messenger
TG·WA·SMS
AI call
in voice
Waitlist
table
01. Loss analysis
First we worked out where the revenue was leaking
We collected the data on no-shows and late cancellations for the period.
We went through the bookings: 22% no-shows, 14% late cancellations.
We measured the lost revenue: about $5.4K a month across the chain.
Bookingsiiko
Analysis
No-shows22%
Cancellations14%
02. Pipeline build
We connected iiko and the channels in 3 weeks
We built a chain of confirmations, reminders and an AI call.
Multi-channel confirmation: Telegram, WhatsApp, SMS.
Reminders 24 hours, 2 hours and 30 minutes before, plus an AI call.
Bookingiiko
n8norchestration
MessengerTG·WA·SMS
AI callin voice
03. Waitlist and dashboard
A freed table does not sit idle
A cancellation goes to the waitlist instantly and the manager sees the forecast.
A table from a cancellation is offered straight from the waitlist.
Dashboard shows confirmations and the occupancy forecast.
Cancellationno-show
Waitlist
Table taken67%
Dashboardforecast
Before / After
What changed for the business
No-shows used to eat revenue and the host called guests by hand. The pipeline removed that.
Before
22% of bookings ended as no-shows
≈$5.4K a month in lost revenue across the chain
3 hours a day of the host calling guests
Only about 50% of guests answered the phone
After
$5.4K/mo of recovered revenue
90% of guests reply in messengers
67% of freed tables get filled again
The AI call and the waitlist run without a host
Paid back in 2.1 months
Outcome
Results in numbers
In 3 months of the pipeline running
$5.4K
a month of recovered revenue
No-shows and idle tables are all but gone
67%
of freed tables get refilled
It was 11%, the waitlist now picks up cancellations
2.1 months
rollout payback
$3.2K rollout against $5.4K a month returned
Let's talk
Want the same result?
Tell us about your project. We reply within one business day and set up a call.