1 · When people actually switch
Almost nobody changes restaurant systems because they found something better. They change because a moment forces it, and there are only four.
- A contract ends — the only planned one, and the only one where you have any leverage
- A site opens — a new room is a clean slate, which is why groups often pilot new systems there
- Something breaks — a terminal dies, a provider closes in your market, or a price change makes the maths impossible
- Somebody leaves — the manager who knew how the old system worked hands in their notice, and the knowledge goes with them
The worst time to choose a system is the week you urgently need one.
Which is the practical argument for looking now rather than later. Not because you should switch today — most people taking a demo are not switching this quarter — but because the four moments above arrive without much notice, and the difference between a considered choice and a panicked one is usually three months of quiet homework.
Avoid December, and avoid your own busiest month. Whatever that is for your room. The best window is a genuinely quiet fortnight where a bad Tuesday is survivable, because there will be one.
2 · What actually breaks
Ranked by how much trouble they cause, rather than by how much attention they get in sales conversations.
| What | Why it hurts | Difficulty |
|---|---|---|
| Staff habit | Muscle memory built over years, undone during service. This is the real cost of switching and it is almost never discussed | Hardest |
| Forward bookings | Reservations already taken for dates after go-live. If these do not move, guests turn up to a diary that has never heard of them | Hard |
| Guest history | Allergies, preferences, visit counts, spend. Rarely portable in any useful form, and the part you cannot rebuild | Hard |
| Card terminals | Contracts and hardware that may be tied to the old provider, sometimes on a longer term than the software | Medium |
| Menu and prices | Tedious rather than difficult, and the point at which someone discovers the printed menu and the till have disagreed for a year | Medium |
| Accounting exports | Your bookkeeper's import routine breaks, usually discovered at month end | Medium |
| Links and QR codes | The booking link on your website, your Google listing, your Instagram bio, and any printed QR still on tables | Easy but forgotten |
The one that catches everyone
Guest history is usually the thing an operator assumes will come across, and it is the thing least likely to. Booking platforms will generally export a list of names and emails; they are far less willing to export visit history, spend and preference notes in a structured form, because that is the asset.
Ask for a sample export before you sign anything with anyone. Not a promise that export is possible — an actual file, with a real week in it, so you can see what shape it arrives in.
3 · What to bring with you
Request all of this from your outgoing provider in writing, early, and before you tell them you are leaving. Some become markedly less helpful afterwards.
- Every forward booking — date, time, party size, name, contact details, and any notes attached
- The full guest list — with visit counts, lifetime spend and dietary notes if they will give them
- Twelve months of trading data — covers and sales by day, so your new system has something to forecast against and you can compare like for like afterwards
- Your menu — with prices, modifiers and any allergen data already recorded
- Twelve months of supplier invoices — for ingredient costs, if you are moving stock control too
- Staff records — names, roles, and hours if the old system held them
Under UK GDPR you have a right to the personal data you are the controller of, in a portable format. Your guests' data is yours, not the platform's, and a provider that makes retrieving it difficult is telling you something useful about how they see the relationship.
4 · The order that works
Roughly a fortnight of elapsed time and considerably less of your own. The sequence matters more than the speed.
Get your data out first
Before anything is configured, before any date is agreed. If the export turns out to be poor, you want to know while you still have every option — including staying put.
Agree the boring things
Table plan, turn times, service charge, deposit policy, cancellation policy, who can void and discount. This is where the time actually goes, and it is worth spending properly because you will be living with it. It is also where you will discover two members of staff have been doing something differently for a year.
Build the menu, then check it against the printed one
Line by line. Every venue that does this finds at least one price that does not match, and it has usually been wrong for months.
Import the diary and the guests
Then open the new system alongside the old one and compare a week. Same bookings, same names, same notes. Anything missing is easier to fix now than on a Friday.
Train on a quiet service, not in a classroom
An hour of real covers teaches more than a morning of demonstration. Run the new till alongside the old one for one service if you can — slower, but nobody is learning under pressure.
Switch the links last
Website, Google, social bios, printed QR codes. Do this only once the diary is live and correct, because it is the moment new bookings start arriving in the new system and the old one stops being the truth.
Keep read-only access to the old system for three months
You will need to look something up. Negotiate this before you give notice — it is usually free and almost never offered.
5 · What to ask any vendor
Including us. If a vendor is evasive on any of these, that is the answer.
Can I see a sample export of my own data, before I sign?
The single most revealing question you can ask. A good answer is a file. A poor answer is a description of a process.
What happens to my data on the day I leave?
You want export available without asking, at no charge, in a machine-readable format, and a stated deletion window afterwards. "Contact your account manager" means the export is a retention conversation.
What is the minimum term, and what is tied to it?
Ask specifically about hardware. A twelve-month software term with a four-year terminal lease is a four-year commitment wearing a disguise.
Who owns the company, and has that changed recently?
The category has consolidated fast. Knowing whether your provider answers to a delivery company, a card issuer or a travel group tells you whose interests shape the roadmap.
What does this cost when I have a good year?
Not the headline rate — the bill at 20% more covers. Per-cover pricing is at its most expensive precisely when you are doing well.
What does it not do?
Every system has gaps. A vendor who cannot name three is either not being straight with you or does not know their own product well enough to be trusted with your service.
6 · Where we would tell you not to bother
This guide is meant to be useful whichever way you go, so it would be odd to end it with a sales pitch. Three cases where changing system is the wrong move.
Your current setup genuinely works
If your food cost is current, your stock take is quick, your invoices are not typed by hand and your bookings do not carry a fee per head, the case for changing is weak. Keep what you have.
A real share of your covers comes from a marketplace network
If people genuinely find you on a platform rather than merely booking through it, that network has value you would be giving up. Check the number in your own reporting before assuming either way — but if it is high, weigh it seriously.
You are in the middle of your busiest quarter
Wait. Whatever the saving, it is not worth a bad December, and any vendor pushing you to switch during your peak is optimising for their quarter rather than your year.
If none of those apply and the four hours a week of admin around your current system are real, then it is worth a conversation — with us or with somebody else.