How to fix a failing restaurant: a 90-day plan

Most failing restaurants are not failing for one reason, and most owners try to fix the wrong thing first. Marketing is the usual instinct: bring more people in and the problem goes away. If the establishment loses money on the average ticket, more people simply loses money faster. The sequence matters more than the effort.

Days 1 to 14 are about cash and facts. Build a thirteen-week cash forecast, list every payable with its due date, and identify the two or three obligations that could close the doors. Talk to landlords and key suppliers early rather than late — terms are almost always negotiable before a missed payment and rarely after. At the same time, pull three months of P&L, a full menu mix report, the labor schedule against actual hours, and count inventory properly. You cannot fix what you have not measured, and almost no failing establishment has a current, trustworthy set of these four.

Days 15 to 30 attack prime cost, because it is the fastest recoverable money. Re-quote the top ten purchase lines. Re-cost every recipe and find the items where the plate cost has drifted past the price. Rebuild the schedule from a forecast with shift-by-shift hour budgets and a sales-per-labor-hour target for each daypart. Three to five points of prime cost is a realistic recovery in a month for an establishment that has never run this discipline, and on $2M of sales that is $60,000 to $100,000 a year.

Days 31 to 50 are the menu. Sort every item by contribution margin in dollars and by units sold. Keep and promote the high-margin sellers, re-engineer the high-margin slow movers with better placement and description, re-price or re-cost the low-margin sellers, and cut the low-margin slow movers outright. A shorter menu also shortens prep, reduces waste, cuts inventory and speeds up service — the second-order savings are usually larger than the pricing gain.

Days 51 to 70 are execution. A shorter menu needs retraining, station rebuilds and updated prep sheets, or the old habits simply return. This is also the point to fix the visible service failures — wait times, table turns, order accuracy — because you are about to spend money bringing people back and you cannot afford them to arrive at the old experience.

Days 71 to 90 are demand. Now the marketing works, because the establishment can absorb it profitably. Start with the cheapest channels that reach people who already know you: the customer list, your local search listing, reviews, and the regulars who quietly stopped coming. Repositioning, price architecture and a proper local campaign follow once the operation holds.

Two decisions cannot be delayed. If the lease is fundamentally unviable — rent above roughly ten percent of realistic sales in most formats — no amount of operational work will save it, and the conversation is with the landlord or about an exit. And if the concept has no trade-area demand, the honest turnaround is a repositioning, not a tune-up. Both are better faced at day ten than at day one hundred.

This is the sequence we run in a turnaround engagement, and the reason our timeline is written as ninety days rather than a promise of overnight results. Stabilize, then repair the margin, then rebuild the demand — in that order, every time.