How to Optimize Your Restaurant Chain in the Age of AI

As restaurant chains’ margins shrink in light of delivery costs and rising food and labor prices, operational leaders are under pressure to find savings through improved efficiency and reduced waste.

Carl Orsbourn of Invisible Technologies says that through careful implementation of AI, food and beverage-based businesses will give themselves the opportunity to substantially improve their bottom line and achieve greater profitability. 

“Waste is a widely understood and serious problem for the restaurant and hospitality sector, and - in an industry where margins can be tight - most businesses know reducing it is one of the real keys to success,” Orsbourn says.   

“Now AI has the potential to revolutionize their operations, analyzing customer orders, inventory levels, and peak dining times to forecast demand and allocate resources with a much higher degree of accuracy.”

“For restaurant chains that recognize this and begin implementing AI capabilities, the result will be that they’re likely to be more profitable and optimized, while those that stick to their old ways will find it increasingly hard to compete.”  

From the Front End to the Back 

Artificial intelligence is already playing a pivotal role in reshaping the restaurant industry when it comes to ordering and delivering food, most notably through third-party delivery apps. These tend to use artificial intelligence to engage customers through a dynamic, personalized experience based on their preferences, buying behavior, and demographic data, as well as to connect them with restaurants and drivers.

But Orsbourn believes this is just the tip of the iceberg, and there is just as much scope for AI to revolutionize restaurant chain operations to generate new efficiencies, cut down on overheads, and significantly reduce waste. 

“Kitchens tend to operate differently from many businesses because there are long periods of quiet followed by bursts of extreme intensity when orders need to be filled. This means they can quickly become overwhelmed in busy periods.”

“You also have lunch or dinner orders arriving from delivery apps at the same time as you’re trying to fill the orders of customers in the restaurant,” he observes. “Given third-party delivery orders now account for around 30% of U.S. restaurant sales, it can force restaurants to operate well beyond capacity in peak periods.”

Orsbourn says that most restaurants attempt to get around this issue by preparing and cooking food ahead of time with the expectation - or at least the hope - that it will be sold: something that can lead to major issues.

“In commercial kitchens, you’ll often hear the phrase ‘get ahead of the wave’,” Orsbourn says. “You know a peak is coming, so you need to prepare and cook as much as possible before it hits.”

“But ultimately, you don’t know when an order will be filled or a driver will arrive to pick up the dish, or even if there will be an order at all. This leads to cold food and a sub-optimal experience for customers. It also leads to the prospect of food needing to be thrown out.”

The peaks and troughs that are an inherent part of the restaurant game also lead to labor inefficiencies, as staff struggle to satisfy both in-house diners and delivery orders simultaneously, increasing the risk of error, as well as the level of demand unpredictability.

A Broader Definition of Waste and Inefficiency 

Orsbourn also observes that the waste and inefficiencies of restaurant chains extend well beyond the volatile nature of their customer flow. For instance, the logistics of delivery means drivers usually have to attend a restaurant in person, physically competing with diners for space and attention. 

“If 30% of restaurant orders are taken up by delivery drivers, they could also be taking up 30% of all space in the parking lot, reducing the potential for dine-in customers to use the restaurant,” he observes.

“Even when the public does get in, they have to vie for the attention of restaurant staff with the delivery guys,   whose capacity for earning is tied to how fast they can deliver. This means they can also tie up human resources, and create a poorer experience for dine-in guests.”  

Then, Orsbourn says, there are the disputed orders where customers claim their meal never arrived. Without proper methods for contesting a claim, sometimes a single missing item means a restaurant will need to refund a customer’s entire check. All up, as much as four percent of revenue is lost through this approach where contesting the charge is not appropriately handled. 

“It’s a shocking number and it means that restaurants can lose as much as half a percentage point of margin to chargebacks alone,” he explains. 

“It happens simply because the fragmented nature of deliveries, with multiple apps and third-party involvement, makes it much more difficult to detect and correct errors and fraud than would be the case for customers eating at the restaurant.”

How AI Brings it Together

While the age of delivery may be presenting fresh challenges for restaurant chains - and exacerbating several existing ones - Orsbourn believes AI will ultimately be used to overcome them. 

“Restaurants can use AI to forecast peak times with a much higher degree of certainty and to resource their inventories and labor in a much more precise way. They can also use it to align order preparation with driver availability so that food is ready just as a driver arrives.”

For this to happen, he says, restaurants need first to harvest their data - something they haven’t traditionally been great at doing. 

“AI relies on data - and lots of it. It’s often something restaurants have in abundance but they don’t utilize it as well as is required.”

Details of transactions, stock levels, customer profiles and order histories, work schedules, operational data, and even external data such as market trends and competitor pricing can all be used to train AI tools in optimizing restaurant operations. 

If gathering and synthesizing all this information seems like a daunting task, Orsbourn says that it shouldn’t be. 

“A third-party provider like Invisible can introduce the technology and human power to collate this data and train AI systems on it in no time at all. In fact, it often takes just a few months to go from not using AI at all, to having it underpin a restaurant chain's operations.

“We act as a seamless bridge to the data that is otherwise disparate and disconnected.”

Because of this, Orsbourn says that most major restaurant chains will be AI-enabled within the next 12-24 months and that those who delay may find themselves falling even further behind.

“The restaurant industry has changed dramatically over the past five years, and this has caused many once profitable businesses to struggle - particularly as delivery takes hold. However, AI capabilities provide a solution to overcome many of the new obstacles holding businesses back.”

“For that reason, successful restaurants in the future will be AI-enabled, while those that stick with the traditional processes and ways of doing business will find it increasingly difficult to compete.” 

“As the Harvard Business Review said in 2023, ‘AI is not going to replace humans, but humans with AI are going to replace humans without AI.’ The same will likely be true for our restaurants.”