Club or restaurant
The bot checks free tables for the time needed and tells the guest your deposit and cancellation terms. It doesn’t take payments. The Tables section shows who is seated, who came without a booking and which booking is next.
Virtual front desk
A Telegram and WhatsApp bot takes bookings while you’re busy with guests
GuestKeepr is a virtual front desk for clubs, restaurants and beauty salons. It answers guests, creates and moves their bookings, and passes the difficult questions on to you. The whole history lives in one dashboard.
- Telegram and WhatsApp
- Russian, English, Romanian, Armenian
- Guest: Good evening! Do you have a table for tomorrow at 21:00? There are six of us23:40
- Bot: Good evening! VIP 1 is free tomorrow at 21:00 for six. What name shall I book it under?23:40
- Guest: Aram, this number23:41
- Bot: Done: VIP 1, tomorrow at 21:00, 6 guests. If your plans change, write here and I’ll move or cancel it.23:41
The screens show a demo venue with made-up guests. The dashboard’s interface is in Russian.
Chapter 1 · An evening without a reply
A guest writes at 23:40. The manager will reply in the morning
Friday, a full house. Scroll through the same evening twice: as it goes now, and as it goes with the bot.
19:30 · Without the bot
A full house, the phone in a pocket
The manager is seating guests. Meanwhile, WhatsApp and Telegram fill up with “Do you have a table for four?”
Conversation (Ani):
- Guest: Good evening! Do you have a table for four at 22:00?
3 chats wait while the manager is on the floor. 3 waiting for a reply.
21:15 · Without the bot
A booking in a private chat
Someone answered on the run and wrote it in the notebook. At the same moment a colleague promised the same table to another guest.
Conversation (Davit):
- Guest: Can we have table 4 at 22:00?
- Staff: Yes, you’re booked 👍
- Note: Table 4 at 22:00 is already promised in another chat
One table, two promises. 5 waiting for a reply.
23:40 · Without the bot
The warmest guest of all
He writes first, he wants to come. But the shift is over, and nobody will answer him until morning.
Conversation (Aram):
- Guest: Do you have a table for tomorrow at 21:00? There are six of us
- Guest: Hello? 🙂
8 chats unanswered by the end of the shift. 8 waiting for a reply.
10:05 · Without the bot
Morning. The reply arrives
The manager works through the chats. Some guests have already settled on another place. What was promised last night the new shift knows only from the notebook.
Conversation (Aram):
- Guest: Do you have a table for tomorrow at 21:00? There are six of us
- Staff: Good morning! Sorry, we’ve only just seen this. Yes, we have a table
- Guest: Thanks, we’ve already booked somewhere else
The guest has gone. Nobody will know how many more there were. 8 waiting for a reply.
Let’s rewind this evening. Now the bot answers
19:30 · With GuestKeepr
A full house, the bot on duty
The manager seats guests while the bot answers in WhatsApp and Telegram: it checks the free tables and asks for a name and a phone number.
Conversation (Ani):
- Guest: Good evening! Do you have a table for four at 22:00?
- Bot: Good evening! Tables 6 and 9 are free at 22:00. Which one shall I book?
An answer at once, and the manager wasn’t pulled away. 0 waiting for a reply.
21:15 · With GuestKeepr
One table, one booking
A table can’t be taken twice for overlapping times: the bot offers a free one. The whole shift sees the booking at once.
Conversation (Davit):
- Guest: Can we have table 4 at 22:00?
- Bot: Table 4 is already taken at 22:00. Table 9 by the window is free. Shall I book it?
- Guest: Yes, please
- Bot: Done: table 9, today at 22:00, 4 guests.
No double booking: the bot never gives away a taken table. 0 waiting for a reply.
23:40 · With GuestKeepr
The shift is over, the bot isn’t
It checks tomorrow’s tables and makes the booking while the guest still wants to come to you, and not elsewhere.
Conversation (Aram):
- Guest: Do you have a table for tomorrow at 21:00? There are six of us
- Bot: VIP 1 is free tomorrow at 21:00 for six. What name shall I book it under?
- Guest: Aram, this number
- Bot: Booked: VIP 1, tomorrow at 21:00, 6 guests.
Booked at night, without the manager. 0 waiting for a reply.
00:20 · With GuestKeepr
A difficult question goes to a person
A corporate party, a special request, a complaint. Whatever is beyond its limits the bot passes to the manager, and it tells the guest so honestly.
Conversation (Lilit):
- Guest: We’d like a corporate party for 80 people on Saturday. Can we have the whole venue?
- Bot: I can’t confirm hiring the whole venue myself. I’ve passed your question to the manager, who will reply in this chat.
The bot called a person instead of making up an answer. 1 waiting for a reply.
10:05 · With GuestKeepr
Morning. It’s all in the dashboard already
The bookings are in the day’s list, regulars are marked, wishes are written down. The log shows who did what: the bot or a member of staff. The corporate party waits for the manager in Conversations.
The new shift knows everything that happened at night. 1 waiting for a reply.
What is lost on evenings like this
Messages get lost between messengers
Guests write in WhatsApp and Telegram at any hour, and the one manager who answers has guests on the floor. Someone is left without a reply and goes elsewhere.
Bookings live in chats and notebooks
Agreed in a private chat, written on a scrap of paper, and in the morning the new shift doesn’t know what was promised. That is how double bookings and forgotten wishes happen.
Nobody remembers what matters most about a guest
That they’re a regular, whether they like a quiet table, what they were unhappy with last time. This memory lives in one employee’s head and leaves with them.
GuestKeepr takes over the messages and gathers everything in one dashboard: bookings, guests, conversations and what the staff did.
Chapter 2 · What it costs
How much you lose and how much you can win back
Put in your own numbers. We count only from what you enter and promise nothing.
Your venue
At night, before opening, or while the manager is on the floor
Answered late or not at all
Usually one in a salon
In the chosen currency
How many of those who book don’t come. The bot doesn’t change this, but a booking won back is worth less
Your assumptions
This is your estimate, not a promise. The starting values are cautious: change them to your own.
Some guests will leave anyway: no places, changed plans
Lost now, a month
1,685€
31.2 bookings a month go unanswered
Can be won back, a month
674€
A year, on your assumptions
8,087€
Lost now, a month: 1,685 €. Can be won back, a month: 674 €. A year, on your assumptions: 8,087 €.
How we counted
- Unanswered bookings: 4 a day × 30% × 26 days = 31.2 a month
- Lost: 31.2 bookings × 3 guests × 20 € × (100% − 10% no-shows) = 1,685 €
- Can be won back: 1,685 € × 40% = 674 € a month
- A year: 674 € × 12 = 8,087 €
This is a calculation from your inputs, not a forecast and not a promise. The cost of GuestKeepr isn’t deducted here: plans haven’t been announced yet.
Chapter 3 · The bot gets more useful
How guests’ questions make the bot more useful
Every conversation is kept in the dashboard. A question on which the bot called a person you see at once. Add the answer to the venue profile, and next time the bot will answer on its own.
The questions and topics are made up. The dashboard has no piles: you notice repeats in Conversations, which has search and an “Awaiting reply” filter.
The week’s questions, by topic
Pick a topic on which the bot still calls a person, and give it an answer.
- 5
Parking
● calls the manager
- “Do you have parking?”
- “Where can I leave the car?”
- “Is the parking paid?”
- +2
- 3
Dress code
● calls the manager
- “What’s your dress code?”
- “Will you let me in wearing trainers?”
- “Can I come in shorts?”
- 3
Opening hours
● calls the manager
- “How late are you open tonight?”
- “What time do you close on Saturday?”
- “Are you open after 2 a.m.?”
- 2
With a dog
● calls the manager
- “Can I come with my dog?”
- “Can I come with a small Pomeranian?”
- 1
Own cake
● calls the manager
- “Can we bring our own cake for a birthday?”
The manager’s queue14
The manager’s queue: 14
- Can I drive right up to the entrance?Parking
- Can I come in shorts?Dress code
- Are you open after 2 a.m.?Opening hours
- Is there anywhere to park nearby?Parking
- Can I come with a small Pomeranian?With a dog
- What time do you close on Saturday?Opening hours
- Is the parking paid?Parking
- Will you let me in wearing trainers?Dress code
- Can we bring our own cake for a birthday?Own cake
- Where can I leave the car?Parking
- How late are you open tonight?Opening hours
- What’s your dress code?Dress code
- Do you have parking?Parking
- Can I come with my dog?With a dog
A new guest asks
- Guest: Can I drive right up to the entrance?
- BotI’ve passed your question to the manager, who will reply right here.
Conversations are collected anonymised
To improve the service, we use conversations without guests’ names and phone numbers. Each venue’s data is isolated from the others. If you don’t want your conversations to take part in improving the service, you can forbid it in the venue settings.
The bot doesn’t become all-knowing
Whatever is beyond its limits it passes to a person. The ready answers to frequent questions are added by you, and the profile holds up to 40 of them. What the bot will learn to do on its own, and when, we don’t promise in advance.
For marketing
What a new guest costs — and where to find more
The host marks where each guest came from, and the dashboard turns these marks into numbers for marketing. We have put them on a page of their own, with calculators that take your numbers.
The cost of a new guest
For each advertising channel: where to add budget and where it goes to waste.
A map of days and hours
When the room is full and when it stands empty.
A promotion for the quiet hours
An example promotion and the arithmetic: what it brings in a month, gifts deducted.
A check a month later
The same numbers before and after the promotion: did it work.
How it works
How the booking bot works: from connecting to the floor
Seven steps on the real screens of the dashboard (its interface is in Russian). The screenshots were taken at a demo club; a salon has the same screens, only with specialists instead of tables.
Staff dashboard
A dashboard for your staff: guests, bookings, log and roles
A reservation book, a record of regulars and the team’s memory in one place, instead of notebooks and private chats.
The guests, bookings and log entries are made up.
Guests and regulars
A card is tied to a phone number: the “regular” mark, favourite table, preferences, staff remarks, visits by day of the week. Duplicates can be merged, and a guest’s data can be erased.
Tap a guest to open their card.
Bookings by day, and their history
Upcoming bookings by day, moving, cancelling, “didn’t come”, history. When you cancel or move a booking, the guest gets a message in their messenger if they have written to the bot before.
Today4 bookings · 13 guests
- 21:30Mariam HovhannisyanregularTable 4 · 3 guestsA table by the window
- 22:00Davit SargsyanTable 9 · 4 guests
- 22:30Aram PoghosyanregularVIP 2 · 4 guestsHis wife’s birthday, a cake without nuts
- 23:30Lilit HarutyunyanTable 2 · 2 guestsAllergic to citrus
The log: who did what
Every action is recorded: who seated guests, moved or cancelled a booking, the bot or a member of staff. No arguments about whose fault it was.
- 21:02Bot (Telegram)Booking createdTable 94 people, Davit Sargsyan · for 22:00
- 21:41LusineGuests seatedTable 43 people, Mariam Hovhannisyan · with a booking · Locals
- 22:05Bot (WhatsApp)Booking movedTable 22 people, Lilit Harutyunyan · to 23:30
- 22:33LusineGuest’s wishes notedVIP 2Aram Poghosyan
- 00:40AnnaTable freedTable 43 people · stayed 2 h 59 min
- 03:30SystemBooking closed: the guests didn’t comeTable 7automatically, 2 hours after its end
Staff and permissions
Everyone sees only what their role needs. Invitations by email, sign-in with Google, sign-in with a PIN on the shared tablet on the floor.
See it through the eyes of
Sees 5 of 11 sections
- Overview
- Tables
- Bookings
- Guests
- Log (hidden)
- Statistics (hidden)
- Conversations
- Settings (hidden)
- Channels (hidden)
- Staff (hidden)
- Roles and permissions (hidden)
These are the default roles: permissions can be changed and you can create roles of your own. The host signs in with a PIN on the shared tablet on the floor; managers and owners don’t sign in with a PIN.
Seating statistics and the cost of a new guest are covered on the page “For marketing”. Open the page →
Works on a phone too
The dashboard opens in a phone’s browser and installs like an app. The host sees the floor and the bookings on a tablet; the manager checks the evening from home.
Who it’s for
For clubs, restaurants and beauty salons
One product; only the words change: a booking or an appointment, a table or a specialist.
Beauty salon
A client writes to the bot, and the bot shows free specialists and times and makes the appointment. It appears in the dashboard at once, and if staff move or cancel it, the client gets a message.
Questions
Frequently asked questions
Can the bot get a booking wrong?
It works only with the guest’s own bookings, on a confirmed number. It always asks for a name, phone number, place and time, and never confirms a booking it didn’t create. A table can’t be double-booked for overlapping times. Mistakes can’t be ruled out entirely, which is why every conversation is saved and visible to you.
What can’t the bot do?
It doesn’t take payments or deposits: it only states the terms as text. It doesn’t remind guests of an upcoming booking. It doesn’t answer the phone. There are no mailings to the guest base and no loyalty programme. There is no public booking widget for websites yet. Questions beyond its limits the bot passes to the manager.
Guests won’t want to talk to a bot
A guest writes where they already talk to you. The bot answers in their language and understands voice messages. A member of staff can take over any conversation and carry on themselves.
Who sees guests’ data?
Each venue’s data is isolated from the others. Staff access is set by roles, and the log shows who did what. The owner can erase all of a guest’s data, and a venue can forbid the use of its anonymised conversations to improve the service.
Is it hard to get started?
Connect Telegram or WhatsApp, add tables or specialists, fill in the venue profile and, if you like, upload the menu. The dashboard works on a phone. We don’t promise how long it takes.
How much does it cost?
Plans and prices haven’t been announced yet.
Will a guest know if I moved or cancelled their booking myself?
Yes, they’ll get a message in their messenger if they have written to the bot before (in WhatsApp the message window is 24 hours). A booking whose guests are already seated can’t be moved or cancelled.
From the blog
All articlesRestaurant no-shows: what to do
Why guests don't turn up for their bookings, how to set rules and keep a record of no-shows. A practical guide.
After-hours booking requests: how not to lose them
Guests write when you're closed. How to organise taking booking requests outside the shift without losing bookings.
A booking bot for a restaurant: what it can do
What a front-desk bot in Telegram and WhatsApp does: it checks places, creates, moves and cancels bookings and calls a person. And what it doesn't do.
How to choose a table booking system
Criteria for a restaurant or a club: channels, roles, data, language and pricing. A checklist that isn't tied to any vendor.
A booking bot for a beauty salon: how it works
A Telegram and WhatsApp bot shows specialists and free times, and the appointment is saved in the dashboard. The scenario, what it can do and its limits.
Booking cancellation policy: how to write one
A template of cancellation rules for a restaurant, club or salon: what to include, how to get it across to the guest, where to keep it.
Try it on your own tables
Connect a channel, add your tables and the venue profile, and the bot will start answering.
Get in touchContact
Ask a question about GuestKeepr, tell us about your venue or ask us to show how it works. We’ll reply in the messenger.