What to order, what to make ahead, when deliveries land, what to move between sites, who is on when - answered per site, from each site’s own history. At one shop, a wrong guess costs an evening. Across a group, it is a budget line.
Every site in the group makes both, every week. The reports you sign price just one of them.
Stock written off at its date. Batches binned at close. Cover on an afternoon the till never justified. This one is measured - it is the waste line and the payroll line - and at group scale it has usually been accepted as the price of doing business.
This one is off the books. The shelf sat empty, the queue gave up, the customer left - and no export recorded any of it. The site reads as slow, so the next order is smaller and the next roster thinner. A mistake nobody measured quietly sets next month’s numbers, site by site.
Not the group average. Each one, from its own history.
Synapse fits one demand forecast per location, from that location’s own history - its peaks, its opening hours, its quiet Tuesdays. The site by the station and the site by the school run different weeks, and a forecast that averages them is wrong at both.
Five decisions read that one fit, each adding what its own job needs. The order, the prep list, the delivery slot, the transfer and the roster agree with each other, because they came from the same picture of that site’s day - and they are taken the same way at every site.
One forecast per site, fitted to that site’s own history. The other five read it.
How much each site buys, for which delivery - and every cut quantity names what stopped it.
What each site makes ahead, window by window, waste weighed against running out.
When each delivery, pickup and handoff should land in a day that is already busy.
When one site will run short and another holds the same units, the group already owns the answer. It names the lane, the day and the quantity - and says when ordering more is cheaper than moving it.
How many people each hour of each site needs, and which legal shifts cover it. A schedule is only called compliant when the records you supply prove it.
Nothing new at the sites. The POS you already run feeds it.
Sales and stock exports from the POS you already run. Synapse reads them as your system writes them - you do not rename a column to fit it, and anything it cannot place is kept, never dropped.
Each site’s order, prep plan, transfers and roster, ready before that site’s order day.
A quantity that was cut names what cut it - shelf life, space, a budget, a supplier maximum. And the same input always produces the same plan, so a plan someone questions is a plan you can replay.
Two of the five decisions - what to buy, and who works - on one made-up restaurant, computed in your browser. Every figure is worked out here, from its data. One site, but the arithmetic each of yours would get.
Hartwell’s is a made-up 60-seat restaurant. Sixteen weeks of its sales are read here, and the week’s order is worked out in your browser. Change how you buy; the order follows.
Loading the demo…
The same sales history, read the other way: who works, when, and what the week’s wages come to.
Loading the demo…
The full suite - the forecast and all five decisions - runs as six demos in the customer portal. Create an account there to run them.
Not a group-wide switch. A trial you can mark.
Pick a handful of sites. Send each one’s sales export and a list of what it stocks.
Synapse is set up against those sites, and their plans start arriving.
Judge the plans the only way that counts: against the orders those sites actually placed. No charge during the pilot.
When the plans hold, the rest of the group is the same step, repeated.
Send a sales export and a list of what you stock, from any site you choose. We set Synapse up against it and run a plan you can check against what that site actually ordered.
Or start in the customer portal: create an account - the demos run there.
Connecting a POS or ERP? Technical reference