Store sales drive the central kitchen, cutting food cost ratio by about 2 percentage points

The company runs several restaurant brands and locations, and has gradually built up centralised purchasing and a central kitchen distribution system.

Stores recorded sales in their POS systems, the central kitchen planned production by experience, and buyers ordered raw materials based on requisition sheets submitted by the stores.

As the number of stores grew, head office realised that although it received sales figures every day, it was hard to determine exactly how much of each ingredient was actually consumed, which stores had higher waste, and how much stock the central kitchen should prepare.

The challenge

Sales data could not be turned directly into ingredient requirements

The POS recorded how many menu items were sold, but there was no consistent data relationship between menus, recipes and ingredients.

Store inventory lacked continuous records

Stores did not record losses, waste, staff meals and ad hoc transfers consistently, so theoretical stock did not match actual quantities on site.

Central kitchen production relied on gut feel

Store orders tended to be either too cautious or too aggressive, leaving some products out of stock while others piled up.

SHEPHARD's digital solution

Step 1: Establish a consistent link between menus and ingredients

Dishes, set meals and retail items sold in stores are linked to standard recipes.

Unit conversion rules and a reasonable loss model allow sales records to be converted into theoretical ingredient consumption.

Step 2: Connect POS and ERP

Store sales, refunds, cancelled orders and promotions are synchronised regularly through an interface.

Business days, stores, sales channels and product codes are mapped consistently.

Step 3: Establish store inventory processes

Store receipts, transfers, losses and stock counts are recorded to a common standard.

Comparing physical counts with theoretical stock reveals abnormal consumption and inventory variances.

Step 4: Build central kitchen production planning

Production suggestions are generated from forecast store demand, sales history, current stock and product shelf life.

The central kitchen schedules production from orders and forecasts, then distributes finished products according to store requirements.

Step 5: Build store performance analytics

Sales mix, theoretical food cost, actual food cost and losses are analysed by brand, store, dish and ingredient.

Stores that consistently deviate from target are flagged for further review.

Implementation approach

Stores with the most complete sales data were chosen as pilots, starting with unified product and recipe data.

Once POS and ERP were integrated, store inventory and central kitchen demand planning were introduced step by step.

For information that cannot be captured automatically, such as on-site losses, recording responsibilities must be clearly assigned; algorithmic estimates cannot replace actual management.

Business value

Head office has moved beyond simply watching revenue to analysing the ingredient consumption, purchasing needs and store operating efficiency behind each sale.

Coordination between the central kitchen and stores has likewise shifted from manual planning to verifiable, data-driven decisions.

Results (approximate): Food cost ratio approx. −2 points; food waste value approx. −25%; store stockout rate approx. −40%.

Metrics tracked on an ongoing basis: Food cost ratio, theoretical-to-actual consumption variance, store stockout rate, food waste value, central kitchen production plan attainment.

What this case teaches us

The heart of digitalisation for restaurant chains is not simply running more stores, but enabling head office to understand the supply chain activity behind every sale.

*To protect client confidentiality, company details in this case study have been anonymised; screenshots are illustrative and results are rounded approximations.*