Six engagements, each re-measured ninety days after handover. Two of them did not hit the agreed number.
Clients are described by type, size and neighborhood rather than by name: an engagement that ends with a menu cut is not something most operators want indexed against their restaurant. Any client is free to identify themselves, and so far none has.
Agreed number: food cost percentage. Result: 34.1% to 29.4% over two quarters.
A menu of forty-one items, of which fourteen sold under six covers a week and between them occupied nine cooler shelves. Removing them released enough prep time to bring pasta extrusion back in house, which is where the margin actually came from. Ticket times at the pass fell four minutes and ten seconds on Friday service. Revenue was flat for six weeks — the owner nearly reversed the cut in week three — then up eleven percent.
Agreed number: cost variance between sites. Result: 11.2 points to 2.6 points.
The same plate cost eleven percent more in one unit than another and nobody could explain it. The cause was three different braise yields recorded nowhere. Shared recipe cards, one weighing protocol, monthly yield checks by the sites themselves. We changed no dish.
Agreed number: prep hours per hundred covers. Result: 18.5 to 14.9.
Breakfast, banqueting and à la carte from one kitchen with three separate par systems and one walk-in. Consolidating pars by ingredient rather than by daypart did most of the work.
Agreed number: ticket time at the pass. Result: 21:40 to 16:05 median Friday.
The range was never the constraint. The expediter was handling both windows and calling times, and the fix was a second call position for four hours on peak nights. Cost: one part-time shift. Saved: about thirty covers a night.
Agreed number: food cost percentage. Result: 32.8% to 31.9%, target was 29%.
We recommended cutting a charcuterie program that was the reason a third of the guests came. The owner was right to refuse, and we should have caught it in the service visit rather than at week five. The engagement fee was reduced by half at our initiative. The lesson — measure why people come before you measure what they cost — changed how we run the first visit.
Agreed number: food cost in the first full quarter. Result: 33.6% against a 30% target.
Opening delayed eleven weeks by a build-out, in which time two purveyor prices moved and the menu we had costed no longer held. We re-costed at no charge; the number still missed. A menu costed twenty weeks before service is a forecast, and we now write the re-cost into the pre-opening contract rather than treating it as goodwill.