For Independent restaurant groups running a few to a dozen or more locations
Catch food cost creep at one location before the monthly P&L does
One location's food cost runs a few points high and it hides inside the group total. It might be portioning, waste, a missed credit, or invoices that never got entered. The P&L shows it weeks later, and nobody remembers what changed.
What it can cost you
Example: a location doing $40,000 a week running 3 points over theoretical food cost = $1,200 a week, or $62,400 a year.An illustration, not a promise. Your Opportunity Scan puts a real number on it.
What the system checks overnight
- Compares actual food cost from invoices and inventory counts to theoretical cost from POS sales and recipes, per location.
- Breaks the gap down by category such as protein, produce, dairy, and dry goods.
- Compares each location to the rest of the group and to its own recent weeks.
- Flags weeks where an inventory count is missing or looks off.
- Flags deliveries that show up in the POS sales period with no matching invoice entered.
Where it reads from
- MarginEdge
- Toast or Square
- QuickBooks Online
What happens when it fires
You talk to the kitchen manager with the specific category and items. The system lists the items driving the gap and drafts a note to the manager asking about portioning, waste, and the missing invoices.
Autonomy level: Suggest only. The brief tells you what changed and what to do. Nothing happens without you.
What we guarantee
Actual versus theoretical food cost recalculated per location each week the invoices and counts are in, with missing counts flagged. We agree on the metric before the Sprint starts and measure it over the first 20 business days after go-live. If we miss it, you get half your Sprint fee back. Dollar savings are tracked in your monthly saves log, reported rather than promised.
Questions
We do not have recipe costing set up for every item. Does this still work?
It works with what you have. Items without recipes are compared on actual cost trends, and the brief tells you which high-selling items still need a recipe.
How often do we need to count inventory?
Weekly counts give the clearest picture. If a location skips a count, the brief says so instead of showing a number it cannot back up.
Do I need new software for this?
No. The system reads from the tools you already use, with read-only access by default. A guardrail layer controls exactly which data the AI can see.
What does it cost?
A 3-week Sprint is $6,000 fixed and includes the morning brief plus one automation tied to an agreed metric. Ongoing support is $1,500 a month, or $500 a month to keep it running without new automations. If the agreed metric is missed over the first 20 business days after go-live, half the Sprint fee is refunded.
How is this different from a dashboard app or a free AI kit?
A dashboard app shows numbers when you go look. A free AI kit gives you the parts, and you do the setup, the upkeep, and the checking. Here the alarms are built for your kind of business, they watch outside data like market rates and rule changes alongside yours, and on Run or Keep-Running, the person who built them keeps them working and logs what they caught.
More alarms for multi-unit restaurant groups
Seven more things worth watching.
- Labor running ahead of salesCompare next week's schedule to forecast sales and reservations per location, and track labor percentage by daypart, so schedules match the business.
- Vendor price increases on key itemsCompare every invoice line on your top items to the last price and the agreed price, across locations, so price creep gets challenged fast.
- Comps and voids anomaliesCompare comps, voids, and discounts per server, manager, and location to their peers and history, so odd patterns surface while they are fresh.
- Review rating dropsRead every new review across your locations each morning, spot repeated complaints, tie them to service days, and have reply drafts ready.
- Reservation no-show patternsTrack no-shows and late cancels by day, time, party size, and booking channel per location, so deposits and overbooking go where they matter.
- Overtime building mid-weekProject who will cross your overtime threshold before the week ends, including staff who work at more than one location, so you can shift hours.
- Menu item margin slippingRecalculate plate cost for every menu item from the latest invoice prices and compare it to menu price, so you reprice or re-spec before margin fades.
See all multi-unit restaurant groups alarms · Browse every industry