Forecasts For Better Food Cost Results

Most restaurants have an automated ordering system (POS) which can be queried for menu item counts for any given data range. This data is quite helpful for the forecasting team in restaurants using a limited menu or a menu with a central focus (either by cuisine or main ingredient). Unwieldy menus do not allow a quick menu item count forecast (using historic POS data) to effectively buy most key items.

At the core, most forecasts begin with the cover count or number of guests expected to arrive in a given time period. Adding the check average to this cover forecast will provide the people designing a shift schedule with a target. Simply take the covers times the check average times the target labor percentage. This is the budget for the period.

Food cost control is not a percentage game in the planning phase. Although most use food cost percentage to measure results, the number won't help you predict what food to order, store, prep and bring to the line for any given period. Cover counts and check averages will provide a broad target. Making the target work is a tougher task.

For many raw ingredients, the ability to use unsold product in future meal periods has a significant impact on your food cost results. Limited menus put a short list of ingredients in play every single day. They don't need to worry how to use any leftover cheese at Domino's Pizza. Pizza is on the menu tomorrow. In contrast, operators using menus relying on daily specials for a major share of sales will have a much more demanding job to use up leftover stock.

The purchasing forecast depends on par stock targets for many shelf stable items. In a typical kitchen, a much greater share of the food cost dollar is spent on perishable meat, fish and produce. The error penalty for over ordering a shelf stable condiment is minimal when compared to fresh white fleshed fish. You will often see seafood specialists with chalkboards. Late in the dinner shift on a busy night, they want to run out of fish if the next day's forecast is for a slow Sunday or Monday.

My favorite Italian restaurant in New York does not offer fish on days when the fish market is closed (Saturday and Sunday). Some chefs pay air freight to get fresh fish in the door as fresh as possible. You want to avoid over buying any highly perishable raw ingredient. Just ask your self a simple question: "If I order more than I can use today, will I be able to sell the leftovers tomorrow?"

Once you move away from fresh fish and poultry, dairy and produce are highly perishable and are frequently purchased by specialists. If your operation doesn't justify hiring a specialist, tracking produce waste can help you adjust your pars quickly. I favor either a two-tier or three-tier par stock model. Order more before busy days and less on slow days.

I believe using sophisticated software (with menu item counts, standard recipes, standard yields and other recipe model factors) are most useful for companies who have never explored these tools. Often, my clients tell me they wish they had analyzed their menu item costs years ago.

The knowledge gained from yield analysis really improves future forecast accuracy. If you learn to translate purchase units of measure into the number of standard portions, you will have a competitive advantage. Your forecast model can now go well beyond covers, check averages and percentages. Take your top 25 items and buy more precisely. The impact is huge.