Restaurants · Hotels · Salons & spas
Guests scan a code at the table and see a menu built around what they actually like, with the dishes you earn most on at the top. You get a dining room you can finally see. And everything the business learns is kept, so it is still there the day you hand it on.
In service from the corner trattoria to the tasting menu Six venues across Bucharest, Lisbon and California.
Elena just scanned.
In service
Fine dining · Corbeanca
Modern bistro · Bucharest
Băneasa · Bucharest
Bistro · Bucharest
Soul food · Inglewood, California
Resto-bar · Bucharest
Most software asks what you use, then hands you a list of the twelve things it talks to. We do it the other way round. There are only three ways to reach a system, and we do all three, so the question is never whether we can connect to yours.
Route one
The clean way. We read and write directly, in real time, the way the vendor intended.
Route two
No API, but it supports delegated access. You grant it, you can revoke it, and we never hold your password.
Route three
Nothing else on offer. An agent works that screen the way a trained employee would, with the same clicks. This is the one nobody else does.
Because of route three, "we do not integrate with that" is not an answer we have to give. If your supplier's system runs in a browser, it is reachable.
your till
Connected Route one · API
Systems named here are examples from this trade, not a list of finished integrations. Which are live for you depends on your setup, and we will tell you exactly where you stand on the first call.
Same six dishes, same kitchen, same prices. The only difference is the order they appear in, and who the list was built for. We print how each number happens, because a percentage with no mechanism beside it is worth nothing.
Average check before
One house list. Every guest sees the same thing, in the same order, whether it is their first visit or their fortieth.
Average check after
Built per guest and weighted to your margins. Read over 1,240 covers at one venue, against its own baseline.A €6.50 lift on 60 covers a night is roughly €142,000 a year
faster table turns
Ordering and paying stop waiting on a waiter, so the queue at the end of the meal disappears and the table frees up sooner.
less spent on admin hours
Taking orders, splitting bills, answering the phone and building next week's rota stop eating paid hours.
more tips for your staff
Guests are asked at the moment they are happiest, and the money goes to the person who served them. Your team notices this in week one.
from yes to live
Menu imported, codes printed, till connected. Nothing bolted to a wall and about ten minutes of briefing for the floor.
Figures from Splitbite venues, each measured against that venue's own baseline before go-live. Your numbers will be your own, and we will show you them the same way.
Not a PDF behind a code. An actual menu that knows who is holding the phone, and is rebuilt for them before it appears. Here is the same list, at the same table, for two different people.
Ranked by what the room orders most. It is the only order available when nobody knows anything about the person sitting down.
Ranked by what she has actually ordered, checked against tonight's margins and what is in the walk-in. Nothing hidden, nothing invented. The same six dishes.
A dish at the top of the list earns nothing if the picture next to it was taken on a phone in bad light. So we come and shoot your food ourselves, during setup and again whenever the menu moves. It is included, not an extra line on the invoice.
Afterwards you can change any of it by asking in plain words. Relight it, clean the plate, straighten the crop. No layers, no masks, no designer, no waiting a week for someone to get back to you.
What your guest sees after scanning the code
Which regular drinks what. Which supplier lets you down in August. How much lamb Easter actually needs. Why the Tuesday special is the Tuesday special. None of it is written anywhere, because you have never needed to write it down.
When a family business changes hands, the next person inherits the building, the name, the staff and the debt. What they do not inherit is thirty years of knowing. They spend their first two years relearning things you already knew, and paying for the lessons twice.
This is the least technical thing we do and the one owners raise with us most often.
of family businesses reach the second generation
The most quoted number in family business, and it has held up for decades.Oregon State University · the 30-13-3 statistic
reach the third
Not usually because the food got worse. Because the thing holding it together was never anywhere but one person's memory.Oregon State University · the 30-13-3 statistic
A by-product of a normal Tuesday.
You are not being asked to document your business. Nobody has time for that, and the projects that try always die by week three. This is the ordinary record of the place running, kept properly instead of thrown away at the end of the night.
The questions a successor cannot ask you.
How much lamb for Easter, and how that has moved over four years. Which sixty guests account for a third of your covers, and what each of them drinks. Which supplier put prices up last spring and what it did to the margin on your best dish. Asked in plain words, answered from your own history.
Even when you are not the one at the door.
The reason a regular feels like a regular is that somebody remembers. Once that lives in the business rather than in one person, a guest gets the same welcome on a night you are not in, from a server hired last month, or from your daughter five years after you stopped.
The one partner who never leaves.
Think of it as a partner who came in on the first day, stood through every service since, and remembers all of it. It does not get tired, take a better offer, or walk out in the middle of August. When you go, it stays, and it briefs whoever comes next.
Not a folder of passwords and a shrug. The same screen you have been using, holding everything the business has learned since the day it went live.
It is also, bluntly, worth money. A buyer paying for a business with a documented guest base and four years of trading history is paying for something they can see, rather than taking your word for it.
Illustrative, to show the shape of it. What you hand over is whatever your own venue has accumulated by then.
A guest's own preferences travel with them, with their permission. So somebody who has never set foot in your place can still walk in and be understood on the first visit.
Elena says she is vegetarian and avoids coriander. She says it once, here.
First visit. The menu already knows, because she agreed to bring her preferences with her.
A stranger to you, and yet not. That is the whole point of the network.
Nearly half of operators are spending more on technology this year, and fewer than one in ten say it has made a real difference. The gap is not ideas. It is that nobody has time to carry them out.Qu · 2026 State of Digital
Nothing here needed you to ask a question or open a report. This is the system noticing something on its own and bringing you a finished thing to approve.
Nothing goes out in your name without you seeing it. You can widen that later for the kinds of campaign you come to trust, and narrow it again whenever you like.
Illustrative, to show the shape of the loop. Real numbers are yours, measured against your own baseline.
The software you keep at the back changes with the trade: a till here, a room system there, a scheduler somewhere else. The guest is the same in all of them.
A real request, answered against real photography
Bookings and in-stay requests, one profile
The same operator view, any trade
The reason hospitality software usually fails is that it asks you to change everything first. This asks you to print some codes and put them on the tables.
Monthly, no lock-in, and your data is yours to take with you. We would rather you stayed because the check went up than because a contract said you had to.
Book the twenty minutesThey do, and they are right to. Surveys keep finding a large majority of diners prefer a printed menu, and one 2024 study put it as high as nine in ten.
That is a fair verdict on what QR codes were used for: saving the cost of printing and the cost of staff, with nothing offered to the guest in return. A static PDF behind a code is worse than paper. The only good reason to put a screen between you and your guest is if the screen knows something paper cannot: who this person is, what they cannot eat, what they loved last time. And anyone who wants a printed menu should still be handed one.
No, and that is deliberate. Rip-and-replace is why most hospitality technology projects stall halfway. Keep your till, your card machine, your payroll and your accountant. We read from them and write back into them, and your staff carry on using the screens they already know.
No. We set it up, we import the menu, we shoot the photography, we connect the systems. What is left for you is approving things you can read in plain English and deciding whether you like them.
If you can read a delivery note and reply to an email, you can run this. The floor briefing is ten minutes because there is almost nothing for staff to learn.
It goes with the business, which is the point. Everything the system has learned since the day you went live is part of what you hand on: the guest profiles, the trading history, the supplier record, every campaign and what it did.
Whoever takes over opens the same screen you have been using and can ask it questions in plain words. They do not spend two years relearning what you already knew. And if you are selling to somebody outside the family, a documented guest base and four years of history is something a buyer can actually inspect, rather than taking your word for it.
The account moves with the ownership. It is your business's record, not ours.
No. Two different things are being mixed up here, and the distinction is the whole design.
Your customer list, meaning who dined with you, when, and what they spent, is yours and stays yours. What travels is the guest's own profile: preferences, allergies and taste, which belong to them, move only if they agree, and can be inspected or deleted by them whenever they like. A competitor never learns that a particular guest visited you. The guest simply does not have to explain themselves twice.
No, and you do not need to hire either. We shoot your dishes during setup and again when the menu changes, and it is included. Afterwards you change images by asking in plain words rather than by opening a design tool.
Campaign artwork works the same way. It comes out in your typefaces and colours by default, so posters and emails look like you without anybody designing them.
Not unless you decide it should. Campaigns are drafted and then wait for you. You see the idea, the artwork and where it is going before anything is published, and you can change or bin any of it.
You can loosen that later for a kind of campaign you have come to trust, and tighten it again at any time. The default is that a human says yes.
Anything touching money, stock or staff waits for your approval before it runs, and every action is logged, reversible and attributed. You decide what it may do alone and what it must bring to you, and you can tighten that at any point.
On allergies specifically: the menu handles dietary preferences, but a guest with a serious allergy should always tell a member of staff. We say that in the app too. Software should not be the last line of defence on something that matters that much.
Yes, and small independents get the most out of it. A chain can already build a picture of a guest across its own sites. A single venue has no way of knowing a first-time guest at all, and no chance of building the thing a chain has. That is exactly the gap this closes.
Twenty minutes, your actual menu, your actual numbers. You will see the guest side, your side, and what it would do in your first week. If it is not obviously worth it for your place, we will tell you.
