Why we started at the other end
Every hospitality platform in Britain starts at the point of sale and works outwards. We started with the half of the day that happens when the doors are shut, and the order you build in shows.
There is a shape nearly every hospitality platform in this country has. It starts at the till. It sells you a terminal, a card machine, a kitchen screen and a printer, and then it adds modules outwards from the counter: online orders, loyalty, bookings, stock, and somewhere near the end of the list, a rota.
It is a sensible shape. The till is where the money is, the hardware is where the contract is, and once a restaurant's sales run through your software, everything else is an upsell. The argument writes itself, and it is a true argument: most places really do end up with four systems that do not speak to each other, and it really does cost them.
We built it in the opposite order, and this is why.
The till is the easy half of the day
Service is the visible part of a restaurant and the part software has already solved. An order goes in, a ticket comes out, a card is tapped, a number lands in a report. The problem is well understood, the data is clean, and twenty companies do it competently.
Now look at the rest of the day.
Somebody opened at six and checked eleven fridges. Somebody counted the walk-in on Sunday night and the number did not match the P&L, and nobody can say whether that is theft, waste, a miscount or a recipe that was costed wrong eighteen months ago. Somebody is building Saturday's rota against a labour budget they are holding in their head. Somebody started last Tuesday and nobody is sure whether they have done their allergen training. The tronc split is in a spreadsheet on one person's laptop.
None of that is clean data. None of it comes out of a till. And all of it is where a restaurant group's margin actually goes.
"Staffing" is one line on somebody else's menu
Open any of the big platforms and look at how they describe the back of house. You will usually find a single bullet: rota, clock-in and permissions.
That line is not wrong. It is just doing an enormous amount of work.
Underneath it sits the rota, the clock-in, the difference between the rostered shift and the actual one, the approval of that difference, the hours flowing into gross pay, the tronc split, the holiday accrual, the leave request that arrived on a Tuesday, the availability that changed, the labour cost against the sales that paid for it, the training record, and the certificate that expires in March.
A company selling to restaurants, salons, shops and plumbers has to average across all four. The rota they can build is the rota all four need. Ours only has to work for a kitchen that changes its mind on a Thursday, and that is the only reason it can go as deep as it does.
What depth actually buys you
Here is the test we use. A feature is deep enough when it knows something the feature next to it recorded.
- The stock count is worth having because the recipe knows what was sold, so the variance has a price on it rather than being a number of missing portions.
- The checklist is worth having because it knows which manager is responsible for that site, so a missed close reaches a person on the day it happened.
- The rota is worth having because the clock-in knows what actually happened, so the hours in payroll are the hours worked rather than the hours planned.
Every one of those sentences has two modules in it. That is what "one app" means, and it is a different claim from "one invoice".
So are you building a till or not?
Yes, and it is being built now — ordering at the table, screens in the kitchen that route by section, and taking a payment where the guest is sitting.
An earlier version of this post was titled "Why Zudaya will never be your till", which was a daft thing to write and is the reason this one has a different name. It confused the order we built things in with a principle. We have never had any objection to the pass; we simply did not start there, because starting there is how the back of house ends up as one bullet point.
Two things follow from that, and they are the useful ones:
Your current EPOS keeps working either way. Sales come in from the system you already have, or by import, and nothing about the rota, the stock count or the checks depends on you replacing your till. If you never want ours, that is a supported way to run Zudaya rather than a grudging one.
The ordering we are building sits on top of the things underneath it. An order placed at a table knows which section makes each line, because the sections are the ones your kitchen actually has rather than three we named. The dish deducts its recipe from stock, because the recipe was already there. A comp has a person's name against it, because the roles were already there. That is the advantage of this order of work, and it is the only advantage — there is nothing clever about being late to a till.
And what we will not pretend
The ordering and payment work above is in progress and not finished. If you are reading this looking for a till today, we do not have one today, and we are not going to put a date on it on a website.
Payroll calculates hours, gross pay and tronc. It does not calculate PAYE or National Insurance and it does not file RTI with HMRC — that needs HMRC-recognised software, and we are not it.
Temperature readings are entered by the person doing the check. Nothing talks to a probe.
And Zudaya is not open yet. There is no list of logos on this site and no testimonials, because there is nobody to put on it. When that changes you will be able to read their names.
If the half of the day after the doors shut is the half that is costing you, tell us what you run and we will show you the parts that apply.