A restaurant can be full every night and still lose money. In the turnarounds we run across Saudi Arabia and the GCC, the cause is rarely the food or the location alone. It is usually a combination of cost leakage, a menu that sells the wrong items, a structure that was built for sales that never came, and delivery growth that adds volume but not margin.
The instinctive response, cutting portions, switching to cheaper ingredients or reducing staff, saves money for a few weeks and then costs guests. The method below finds the money that guests never see.
The method at a glance
- Read the P&L by outlet, channel and item
- Measure the gap between theoretical and actual food cost
- Re-engineer the menu by profit and popularity
- Fix purchasing: specs, suppliers and yields
- Control waste and inventory
- Schedule labour to the sales curve
- Calculate the true margin of delivery
- Review rent, and centralize production where it pays
1. Read the P&L the right way
A single consolidated P&L hides the problem. Split it three ways before you change anything: by outlet (which sites make money), by channel (dine-in, takeaway, delivery, catering) and by item (from your POS data). Then calculate a small set of ratios on net sales, excluding VAT:
| Ratio | How to calculate it | What it tells you |
|---|---|---|
| Food and beverage cost % | Cost of goods sold ÷ net sales | Whether recipes, pricing and purchasing are under control |
| Labour cost % | All staff costs, including housing, visas, GOSI and benefits ÷ net sales | Whether the team structure fits the sales level |
| Prime cost % | (Cost of goods sold + labour cost) ÷ net sales | The best single indicator of operational health |
| Occupancy cost % | Rent, service charges and related fees ÷ net sales | Whether the site can carry its rent |
| Contribution by outlet | Outlet sales − outlet direct costs | Which sites fund the business and which ones drain it |
As a rule of thumb, many operators aim to keep prime cost at around 60% of sales or below and occupancy cost at roughly 10–15%. The right targets depend on the concept and the price point, but a business well above both is unlikely to be fixed by sales growth alone.
Cost of goods sold must come from stock counts, not from purchases: opening stock + purchases − closing stock. Without a reliable month-end count, the food cost on your P&L is a guess.
2. Find the gap between theoretical and actual food cost
Theoretical food cost is what your sales should have cost if every dish was made exactly to recipe: the sum of each item’s plate cost multiplied by the number sold. Actual food cost is what you really used, from the stock count. The difference is the variance, and it is usually the fastest money in a turnaround.
A variance of several percentage points of sales points to over-portioning, unrecorded waste, staff meals, missing stock, recipes that changed in the kitchen but not on paper, or supplier prices that crept up. To close it you need three things: a costed standard recipe for every item, portion tools in the kitchen (scales, scoops, ladles, plating photos), and a weekly count of your highest-value items.
This is the most important point of the whole guide: closing the variance does not change the dish. The guest gets exactly the plate the recipe describes. You simply stop paying for the food that never reached the plate.
3. Re-engineer the menu by profit and popularity
Menu engineering ranks every item on two measures: contribution margin (menu price excluding VAT minus plate cost) and popularity (its share of items sold in its category). Each item then falls into one of four groups:
| Group | Profit / popularity | Typical action |
|---|---|---|
| Stars | High margin, high popularity | Protect the recipe and quality; give them the best menu position |
| Plowhorses | Low margin, high popularity | Review the recipe cost and the price carefully; adjust the garnish or side, not the core |
| Puzzles | High margin, low popularity | Reposition, rename, recommend through staff, or rework |
| Dogs | Low margin, low popularity | Remove, unless they serve a clear purpose for a group of guests |
A shorter menu usually costs less to run: fewer ingredients to buy and store, less prep, less waste and faster service. In many turnarounds, removing slow items does more for profit than any price increase.
Pricing should follow the analysis, not replace it. Raise prices first on items where you have a clear advantage, and check every change against competitors in your own area and price band.
4. Fix purchasing: specs, suppliers and yields
Write a specification for every key ingredient (size, grade, origin, pack, shelf life) so suppliers quote on the same product and the kitchen receives the same product every time. Then:
- Consolidate suppliers and volumes, and renegotiate prices, payment terms and delivery days
- Check every invoice against the agreed price list; price creep is common and rarely noticed
- Buy on usable cost, not on price per kilo: usable cost = purchase price ÷ yield %. A cheaper product with a lower yield can cost more on the plate
- Test substitutes only where guests cannot taste the difference, and prove it with a blind tasting
5. Control waste and inventory
Waste is bought, stored, prepared and then thrown away, so it carries the full cost of the ingredient plus the labour. Control it with prep sheets based on a sales forecast, par levels for each station, first-in-first-out storage with clear labels and dates, and a daily waste log that the head chef reviews. What gets recorded gets reduced.
6. Schedule labour to the sales curve
Labour is often the largest cost after food, and the most poorly planned. Use your POS data to build an hourly sales curve for each day of the week, then roster to it. Measure sales per labour hour and compare shifts, not just months.
The goal is not fewer people at peak times. It is fewer idle hours in quiet ones: staggered start times, split shifts where allowed, cross-trained staff who can move between stations, and prep moved into quiet periods. In Saudi Arabia, plan the structure together with your Saudization (Nitaqat) targets so the savings do not create a compliance problem.
7. Calculate the true margin of delivery
Delivery can grow sales and shrink profit at the same time. For each platform, calculate the margin per order:
Delivery margin = menu price − platform commission − packaging − promotions you fund − food cost
Use the commission rates in your own contracts. If the margin is too thin, build a delivery menu: items that travel well, are quick to prepare and are priced for the channel, with packaging costed into the price. Decide which promotions you fund and which ones you decline.
8. Review rent, and centralize where it pays
If an outlet cannot carry its occupancy cost at realistic sales, no amount of kitchen efficiency will save it. Review each lease: rent, service charges, escalations, turnover rent and exit clauses. The options are to renegotiate, to change the format to fit the space, to relocate, or to exit. Sometimes closing one outlet is what saves the rest.
For groups with several outlets or brands, a central kitchen or bakery can reduce labour, waste and variation for the items that suit batch production, such as sauces, doughs, pastries and marinated proteins. It only pays when volumes are high enough and the logistics are planned, so model it before you build it.
Quality guardrails: how to cut cost safely
Every change in a turnaround follows the same rules, so the savings never reach the guest:
- Lock the signature dishes and the core guest experience before changing anything
- Pass every recipe or supplier change through a blind tasting against the current version
- Track guest scores, reviews and complaints weekly, next to the savings
- Change portions only where they are inconsistent, never to shrink the plate
- Train the team on every new standard, so results last
A 90-day turnaround plan
| Period | Focus |
|---|---|
| Days 1–14 | Profit diagnostic: P&L by outlet, channel and item, recipe costing, food cost variance, menu engineering, purchasing, labour, delivery and lease review. Output: a savings plan with a value for every action. |
| Days 15–45 | Quick wins: portion control, invoice checks, waste logs, rosters built on the sales curve, removal of the weakest menu items. |
| Days 46–90 | Structural changes: menu re-engineering and pricing, supplier contracts, delivery menu, lease negotiations, SOPs and training. |
| Every week | A short KPI review: sales, prime cost, food cost variance, waste, labour hours and guest scores. |
Expand only what already works
A second branch multiplies everything the first one does, including its problems. Before expanding, a concept should pass five tests: the existing unit is profitable after rent and full labour; recipes, SOPs and training are written down and work without the founder on the floor; there is a manager ready to run the new site; the supply chain can serve more outlets at the same quality; and you know which locations, sizes and rents the concept needs to break even.
Once those tests are passed, the next step is choosing the right format for each site (full restaurant, express, kiosk or cloud kitchen) and, when the system is proven, preparing for franchising.
This is the method behind our Profit Turnaround programme. For a four-brand F&B group with 300 to 400 employees, it cut costs by more than 30% and returned the group to profit without relying on sales growth. Read the case study.
The ratios and rules of thumb in this guide are general benchmarks, not targets for every business. The right numbers depend on your concept, location, price point and channel mix.