Nowadays, quick-service and fast-casual chains are under especially heavy amounts of pressure when it comes to meeting customer demand for quantity without sacrificing quality. Thanks to tech, managing — and predicting — that demand is getting a little easier.
Case in point: Houston-based enterprise tech consulting firm Smartbridge launched KitchIntel, a software solution for restaurants to more precisely monitor their proteins, in 2019. Smartbridge’s Marketing Manager Brooke Browne and the company’s Director of Digital Innovations Deepthi Raju got on the phone this week to talk me through how the system works.
Backing up for a second, Browne and Raju told me the company has chosen to focus on protein for now because chicken, beef, and other meats form the base of so many dishes on QSR menus. As well, most proteins spoil easily and have to be cooked in certain ways to ensure food safety.
The KitchIntel software takes a lot of the guesswork out of the cooking and ordering processes by telling cooks how much food to cook when and also sends them alerts at key steps in the cooking process. A dashboard interface displays which proteins are currently on the grill, how long they have been cooking, what to cook next, and even when to flip the meat over.
All of that information comes from data inputted in the system. In the case of how to cook the proteins, that information would come from corporate, based on company recipes. “The beauty of this application is that it has the benefit of standardization as to how [restaurant chains] cook their main proteins,” Raju said on the call. “It reduces the amount of time the manager has to spend on training the new employee.”
KitchIntel is at this point most useful to quick-service and fast-casual restaurant chains that have to cook said proteins in huge batches over an extended period of time in order to meet customer demand. Think restaurants that specialize in slow-cooked chicken dishes or a chain like Chipotle that has to keep large quantities of protein on hand around the clock. At the moment, the company is more focused on these types of chains than one like, say, McDonald’s, whose operations more involve making single patties on the grill.
Perhaps even more significantly, a system like KitchIntel’s can help kitchens predict how much food they need to cook in the first place. The system integrates with other pieces of the restaurant’s technology stack, which means it can pull historical sales data from the POS to predict demand on any given day for any given shift. By way of example, Browne referenced one QSR client during our talk (name withheld) whose menu heavily features grilled chicken. After rolling the KitchIntel system out to over 130 of its locations, the chain has reported food cooked closer to its time of sale, ensuring fresher food for the customer and less food waste thanks to better communication with inventory systems and more accurate predictions of how much to cook.
Browne says the company also offers an AI component that can provide even more granular data than historical sales. As we’ve seen with companies like McDonald’s, who implemented Dynamic Yield’s AI tech into its drive thrus last year, using AI and machine learning to factor in weather and traffic data for each individual store can help restaurant chains make even more accurate predictions in terms of how many customers to expect. Five years ago, rainy weather might have meant a store didn’t prep as much chicken because it anticipated that most people who stay in. Nowadays, though, delivery orders often spike during rainy weather. AI that ingests weather data and feeds it into the KitchIntel system of a restaurant could help restaurants know exactly how much chicken (or beef or pork) to prep in order to better meet forthcoming demand.
“A lot of decisions are made just by looking into the dining room,” Raju said. “That throws them off completely when there is a weather situation. While the dining room may look empty, it’s the drive-thru or the Grubhub people that are getting busy. So really incorporating AI into the sales prediction aspect of it is one thing that makes the solution unique.”
As demand for off-premises orders — delivery, takeout, drive-thru, etc. — drives the bulk of restaurant sales over the next decade, QSRs and fast-casual restaurants in particular will find themselves increasingly under pressure to deliver high volumes of protein-based meals to customers without sacrificing speed or quality. As in most parts of the restaurant industry these days, especially when it comes to behind-the-scenes operations, more data for smarter, more accurate predictions seems to be the key.
