How Technology Moves You and Your Customer

Answers to Rising Food Costs

Answers to Rising Food Costs

Published in the July 2014 issue

Caught in the vise of rising food costs and consumer price resistance, restaurant operators should welcome all and any new ways to loosen the screws and reclaim profits. Or so you’d think. One countermeasure is comparison sourcing, but purchasers have little time for it.


e*Restaurant’s Universal Forecasting engine, based on sales history, drives the output of suggested ordering, prep-and-pull, and employee scheduling. It factors in predicted weather affects and allows managers to also accept the historical affects of events such as breakfasts and parades. Suppose a GM observes, say, that rain isn’t affecting sales this year? She can dial/adjust the base forecast up or down.

That’s a problem being tackled by a number of new online market platforms. Foodem, based in Baltimore and active among Washington D.C. restaurants, runs a site that connects food distributors and farmers with wholesale food buyers, who can compare prices and look up user ratings. Standing orders can be stored on the system and a regional concentration helps lower costs. Another virtual market in this category is DineMarket and local to the NYC area is FarmersWeb. itradenetwork.com, a foodservice trading network, works more actively with restaurants and chains to analyze the supply chain, contracts, and manage the entire transaction,

Another counter measure? Getting inventory right. More than almost any other business, restaurants are run on instinct, especially when it comes to food orders. But knowing what food items to order, in what quantities and when, that is something inventory and purchasing algorithms do much better than human instinct.

Here’s just one reason why restaurant veterans should trust a restaurant management program instead of instinct when it comes to ordering: Instinct can’t remember and analyze the last 12 years of sales, or know how a ten-degree temperature change impacted dessert sales.

But Altametrics’ e*Restaurant package, for example, does. On the supply chain and inventory management sides of its Universal Forecasting engine, it factors in such data as weather predictions to arrive at purchase predictions and suggested orders. Interestingly, the weather forecast doesn’t have to turn out to be correct for the purchasing prediction to be proven right. For a full-service restaurant, for example, widespread blizzard predictions will deliver a flurry of empty tables, even if the promised blizzard never materializes. For a QSR, on the other hand, customers act more spontaneously, so a no-show snowstorm won’t kill business. The software knows the difference. The software even considers whether the restaurant is deep inside a mall (increased snow business) or freestanding (decreased snow business).


The Suggested Ordering function of e*Restaurant’s Inventory application analyzes historical usage to come up with orders. Managers need only input quantities and items they wish to change. The app integrates with major distributors’ ordering systems, to reduce purchasing to a few clicks.

It also allows for a wide range of local events—such as school sporting events or municipal parades—to be programmed into the system, where historical data on those events can be analyzed, said Gurminder Dhamija, Altametrics’ director of professional services and engineering.

The psychological problem, says Dan DeMille, a client services manager with Altametrics, is something called fear of the bus. “Operators always to have 10 cases on hand, just in case” a busload of people shows up. “So they continue to order the same quantities they have always have.”

Overdoing food stock obviously not only can lead to spoiled food, but can it actually fuel theft: DeMille argues that having ten cases of steaks in your freezers increases theft simply because it’s less immediately obvious if one goes missing.

Make it a contest — with consequences
Perhaps the best way to persuade people to use the software is to tie bonuses into setting proper stock levels, DeMille suggests. Managers will be given the option to override the software recommendations, but if they do such an override and they don’t outperform the software’s recommendations, their bonus gets cut.  Quickly, they realize that emotionless analytics are the much safer route to a full bonus.

The software has a tamper-resistant element that prevents managers from playing with the inventory reports to make their forecasts look more accurate. “If you’re on an older system, you could repeatedly play with your numbers,” says Al Schave, Altametrics solutions consultant. The bosses “then don’t know how many times you changed your orders of chicken. With our software, someone will be able to see that a manager went in and played with it.”

Ordering levels also get out of whack when managers are swapped in and out of new locations. It’s a good strategy for getting a fresh perspective, but it also tends to rob a restaurant of the benefits of experience with its particular clientele. But accumulate that clientele’s history by loading it from POS into software, and that problem is avoided.

eRestaurantTablet
Inventory on the iPad reduces a two-step process to one in-stockroom step, by replacing paper and keypad entry. e*Restaurant accepts multiple amounts of the same item in different units of measure, so that all ground beef, for example, gets counted whether in pounds or patties.

Another problem spot: missed handoffs between shift managers. “Because you work different shifts, you don’t know what the other manager did or failed to do,” Schave says. “That can create double orders or missed orders. The software keeps an action items list of what still needs to be accomplished.”

Data entry once — at the inventory site
While inventory software concerns itself with proper order levels, it can also speed up the stock-counting process itself. Simply by replacing paper inventory forms (which must then be keyed in manually — a second step) with mobile devices, managers can shave hours off a typical five-hour task and improve accuracy to boot. Data is entered only once, in the stockroom, with the tablet.

Attach a scanner and the process goes much faster. If the software supports item-level RFID as well as bar-code scanning, as e*Restaurant does, a large number of items can be scanned simultaneously at longer range and without clear lines of sight, potentially accelerating the process by an order of magnitude.

“You leverage the mobile devices that your managers already have, and you take them where the boxes are kept. In this way, you don’t have to print anything out,” Dhamija points out. The e*Restaurant mobile app also sports a simplified graphical user interface (GUI), designed throughout to leverage a mobile device’s touch-screen and other capabilities.

By forecast; not by feel
We’ve noted above that forecasting software can spot subtle trends—associated with unusual weather or days of the week or certain months—that a manager might not see. Even more likely, though, are trends that management might see but unintentionally exaggerate. If five customers engage in a specific purchase behavior, a manager might extrapolate that data more than reality supports. The software doesn’t.

(Note: a related Altametrics app, the Xformity business intelligence suite,  takes extrapolation further, revealing correlations and irregularities that would otherwise go undetected. More on this later this year.)

eRestaurantPhone
The smartphone version of of e*Restaurant Inventory shows one item at a time, with touch-sliders or taps to enter quantities.

e*Restaurant also allows for extremely local variables to be entered and then analyzed. Every time there is construction on a nearby highway, for example, how does it impact business? That roadwork might be diverting customers from specific office complexes, who tend to make similar lunch-hour decisions. It would be hard for a manager to know that a blockage on Exit 63 of Route 80 cuts down on salad purchases, but the software could, if the event was noted in the system.

Another example: What is the impact if a key neighboring business—perhaps a movie theater or bookstore—closes? Inversely, if that neighbor changes its hours or days open, how should your inventory change?

The key to saving on food costs is attacking that profit-eating variance between theoretical food costs (how much your team thinks it needs to buy) and actual costs (what customers actually buy, plus what gets thrown out or walks away). What looks to humans like haphazard and irrational buying patterns are typically quite predictable and consistent to forecasting software, which avoids emotion, fears no bus, and simply examines huge amounts of historical data. It will spot patterns that a manager—dealing with 18 people with 18 different requests and/or complaints—will almost certainly miss.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

  • Also In This Issue

    \"Digital

    Digital Menus

    \"Editor's

    Editor's Choice

    \"Google

    Google Glass for Chefs