Want to move your company forward faster? Smarter? Consider the many ways to tap into the power of artificial intelligence. The technology has advanced to the point that business leaders need to take a hard look at how AI can benefit their companies, said Jim Lecinski, clinical associate professor of marketing at Northwestern University’s Kellogg School of Management. “Businesses big and small can take advantage of this machine learning to save money and be more profitable,” he said.

Essentially, it’s time to get in the game. Parker’s convenience stores did just that. Both food and technology are seamlessly integrated into the DNA at Parker’s Kitchen, a fast-growing brand under the Parker’s umbrella. The 68-store chain based in Savannah, Georgia, operates 48 Parker’s Kitchens today, all of which utilize “smart kitchen” technology. A company-wide rollout took place in 2020.

An early adopter of tech-based components, Parker’s relies on its in-house internal innovation and digital transformation team to build its tech systems from scratch with a focus on AI-based predictive analytics to optimize store-level performance. AI touches the company's food offering, self-checkout lanes and rewards app, to name a few. “We have a phenomenal innovation team,” said Heather Davis, director of foodservice at Parker’s. “We dream it, they create it.”

While the chain is likely leap years ahead of the norm in applying smart technology, Davis does remember when operations were far less tech sophisticated. “When we were first looking at the smart kitchen concept, everything functioned on a completely manual process, pulling historical sales and using averages,” she said. “I had a huge Excel document with a tab for every store that I manually updated. Then at the store level, hoped they printed the right sheets.” Deciding quantities to prepare for the food bar were basically guideposts from an individual, she said, largely determined by the store manager’s gut instinct.

That all burst wide open when Parker’s initiated its AI-powered smart kitchen. “We now know that a certain location will need X number of potato logs Monday at noon. Stores can produce just-in-time food, and the store level doesn’t have to retain that information. The screen is their guide,” she said.

The immense amount of data has allowed for a more precise mapping out of what staff will cook. “We have adjusted our daily flow and production schedules based on when stores are cooking the most food. We continue to make adjustments in expectations,” said Davis. 

NOW IN SESSION

While Parker’s fine-tunes its AI components, other convenience retailers still may be trying to grasp exactly how AI works. Understanding the path and pathway from the past to today’s version of AI can help business leaders unlock its potential for the future. Lecinski explores the past, present and future of AI in his new book “The AI Marketing Canvas.” He also shares his expertise each year as the program director of the NACS Marketing Leadership Program at Kellogg.

AI’s origins date back to World War II, an important notation Lecinski said, because that is when people started to explore how machines could think beyond what humans exclusively told them to do. During that war, an early computer pioneer named Alan Turing created a device to unscramble Nazi codes and help the British Navy avoid submarines. His early work is credited with paving the way for what we now call artificial intelligence, a term that was coined in the 1950s.

Fast forward to 2022 and AI is part of nearly every business conversation, from forecasting to marketing to finance. Why the long lag time in mainstreaming AI? The short answer is a common one in the technology field: “The problem was that it was really expensive and not enough people knew how to do it back then,” explained Lecinski. “Today, your cellphone is millions of times more powerful than the rocket NASA used to land on the moon.”

AI is important and relevant to the c-store industry because it is a new way to  apply practical computer science.

AI is a broad subject matter, said Lecinski, much like retail itself. Many segments exist under that big umbrella of “retail,” he pointed out, from quick-serve restaurants to high-end clothing stores to c-stores and more. The same is true with AI, which he defines as the study or discipline of how computer systems can act in a humanlike way. Subsets he points to under that umbrella include robotics, fraud detection and marketing, among others.

“AI is important and relevant to the c-store industry because it is a new way to apply practical computer science,” explained Lecinski. The prior approach was to tell computers what to do in a language the computer understood. For example, if the numbers in column X are less than 10, do this. Humans would program in every possible move. As an example, he points to IBM’s famous Deep Blue chess-playing supercomputer, which in 1985 became the first computer to win a chess match against a world champion player. It serves as a classic man vs. machine lesson that demonstrated to the public that computers could handle complex calculations. Deep Blue was programmed for every possible chess move on the board. AI today takes a more advanced approach. Instead of programming if/then scenarios for every eventuality, computers are given different sets of inputs and shown the outputs. “Essentially, with AI, what we are telling computers is ‘you figure it out,’” Lecinski said.

EYES FORWARD 
Lecinski walked through an example of AI using credit card fraud. “Let’s say I show the computer a million credit card transactions that I know are fraudulent. Then I give the computer another million transactions and tell the computer to predict which are fraudulent. There are no if/then components, I just say ‘you figure it out.’” That is AI in a nutshell, he said, and it “has unlocked a whole new way of using computers.” And the computer keeps learning. “What the machine predicts, I go back in and tell the machine what is right and what is wrong, and it continues to learn.” In essence, it’s about doing an increasingly better job of predicting. “The learning loop gets better each time.”

AI continues to move forward with machine learning, said Gray Taylor, executive director of Conexxus, a nonprofit technology organization focused on tech standards and related advocacy in the convenience and fuel retailing industry. “It keeps building and building. It’s a self-perpetuating thing that keeps asking ‘why?’” he said.

In the future, Taylor foresees AI playing a key role in refining forecasts in a lot of different areas, such as the supply chain. For example, he pointed to the aluminum can shortage that hit during the pandemic. “Manufacturers could not get enough aluminum. [As an industry], we were not forecasting out far enough for them,” he said. “Machine learning will change that. It will consolidate information and provide better inputs 90 days out.” The idea is to give the supply chain enough time to find alternative sources, Taylor noted, adding that while AI will make that happen, it’s not quite there yet. “We are working on adopting that; we are working on the deep learning.”

WAYS TO JUMP IN 

Companies shy away from AI because of a perceived large financial investment. “It does not have to be a big investment early on,” Lecinski stressed. To overcome the hurdle, he steers companies to adopt more of a “crawl, walk, run” approach. His next piece of advice involves a low-cost approach to entry. “Go work with your existing partners,” he said. “Ask them, ‘What machine learning options do you have that we can try ... for no money?’”

Manufacturers like The Hershey Company and PepsiCo Inc. routinely make news in their use of AI. “While AI is getting applied at the manufacturer level first, we are increasingly seeing retailers work on machine learning,” said Taylor.

The forecasting pieces are a no-brainer in terms of what to add. There are levels of all this technology now. It’s about finding a good partner to work with you.

Al Indig, senior engineer – project manager at New York-based PreciTaste, believes machine learning/AI “can be a huge help in convenience stores, just like it is helping QSRs—essentially allowing fewer or even just one person to manage food efficiently.” This higher level of technology is being applied at larger QSRs now, he said, noting that it’s not uncommon for bigger chains to lead with technology. PreciTaste’s platform centers around AI and machine learning in professional kitchens.

One conversation Indig often has with QSR chain leaders centers on the need to drive efficiency in a short-staffed environment—clearly a common scenario for food-focused c-stores. How the AI processing model helps, Indig explained, is by taking vision images and using an array of data points, which are then fed into the computer, to guide employees. In a practical sense, the AI agent makes decisions to assist the crew. For example, he said, it would tell employees when hot food was sitting out too long, eliminating both the need for an employee to set any kind of timer system and reducing any training hiccups with a new employee. “The computer knows all the rules; the staff just need to follow the computer prompts,” he said. “The technology doesn’t replace people, but it does help when short-staffed. One of the biggest wins is that it optimizes operations in the kitchen at that moment.”

Parker’s looks everywhere it can to optimize kitchen operations and had just that in mind when the team attended the NACS Show in October 2021. They were specifically homing in on equipment partners with telemetry, Davis said, “to see what can push and pull data into another system.” The premise is for Parker’s innovation team to be able to “talk tech” with the equipment manufacturer’s tech team to gain access to their application programming interface. Essentially that would allow Parker’s to access all the raw data and build it into their system, whether reporting or smart kitchen forecasting.

While AI is getting applied at the manufacturer level first, we are increasingly seeing retailers work on machine learning.

The end goal for Parker’s would be to tie everything together in a fully integrated ecosystem versus having different systems for different vendor components, for example ovens and printers. Davis described what this looks like at the store level: “We want people to hit the screen to show they cooked chicken tenders and just print from that screen, eliminating a step to have to go to the printer and do it there.”

For anyone considering chasing AI and related technologies, Davis suggests examining what exists in the marketplace today if a company isn’t ready to build a proprietary system. “Identify existing partnerships and ask what they can offer to achieve some of these things. It’s out there now,” she said. “The forecasting pieces are a no-brainer in terms of what to add. There are levels of all this technology now. It’s about finding a good partner to work with you.”

The Parker’s team continues to look for ways to maximize data, but Davis added that the first round was not perfect. “We had to beat it down before we rolled it out and continue to make adjustments. Now, people get excited about this.”

Whatever stage a company’s learnings are around AI, note that 2022 marks the year that most experts consider it a mature enough field to jump into. According to the most recent findings from the McKinsey Global Survey on AI, published in November 2020, increasingly, organizations using AI as a tool for generating value are finding that value in the form of revenue. And, according to the report, those companies plan to accelerate their investment in AI.