Are you using artificial intelligence (AI) in your convenience-store menu development?
“AI is not just a buzzword in menu development and in foodservice,” said Kyle Drenon, pictured with Supper Co. co-owner Kesha Alexander at CSP’s C-Store Foodservice Forum on June 17. “It is the new normal. It is the future, whether we like it or not.”
Supper Co. is a Springfield, Missouri-based ad agency.
AI can help retailers spot cravings before they happen, and Drenon took a deep dive into Seattle-based Starbucks’ AI, Deep Brew.
‘Drinks like this are the result of the analysis that the AI did.’
“That is their proprietary AI analysis tool that they built in 2019 where they were doing AI before AI was cool,” Drenon said.
Deep Brew was instrumental in bringing new menu items to Starbucks’ customers, like the Iced Brown Sugar Oatmilk Shaken Espresso, he said.
“Drinks like this are the result of the analysis that the AI did,” he said.
From 2021 to 2023, cold beverages surged on the Starbucks menu, said Drenon, who added that Deep Brew analyzing seasonal and regional trends across all Starbucks transactions “detected that trend before it happened,” he said.
Analysis discovered items people were adding to their drinks, like cold foam, Drenon said.
“They saw an increase across seasons of iced espresso drinks starting to pick up in all these different regions, and they were like, ‘That's a trend. We’re gonna be ready for it.’ And they were,” he said.
Drenon said that 76% of all of Starbucks sales now are cold drinks—“76% after decades of hot coffee dominance—and they were ready for it, and they had a 9% same-store sales growth because they predicted that that was going to happen. They had a killer 2021, and it was because they were ready for this massive trend.”
How C-Stores Can Benefit
C-store retailers can emulate what Starbucks did with their own AI systems, Drenon said, adding that if c-stores don’t have an IT system yet, they will soon.
“If you want to do this stuff, talk to your IT team before you do it,” he added.
To get started, isolate a sales period, said Drenon, who recommended three years of sales data. Retailers should download from their point-of-sale system into a spreadsheet as much data as possible, including dates, locations, customizations and full basket size.
“Then upload that into your AI and begin talking to it like you're having a conversation,” Drenon said. “Because that’s what AI can do for you. And that’s essentially what the Deep Brew technology was: A conversation with the data. And your POS probably has ways to measure trends and things like that, but it wouldn't be as successful as what you can do with AI.”
He said a conversation can include asking: What drink and menu item combinations are popular in the state of Missouri in the month of June? “And it will give you the answer based on the data that you’ve uploaded,” Drenon said. “You can say, ‘What things have been trending upward over the last three years for late-night snacks in California?’”
‘Keep Talking’
Once a retailer has those trends analyzed, they should keep talking to the AI until they start to see big differences. “Like this was a huge growth,” Drenon said. “You’re gonna detect those things in your data.”
A retailer can then ask AI to create LTOs based on what they already have in their store.
“If you do chicken, make it about chicken,” Drenon said. “What are the most popular drinks that go with chicken? Let’s say it’s Dr Pepper, and we’re gonna do a snack wrap with a dirty soda Dr Pepper. And that’s an LTO. And that can come from an active brainstorm with AI that you can do right now.”
Retailers also can also ask for suggestions based on trends in the restaurant industry and their competitors.
“Just keep asking—and ask the AI, 'What are my competitors doing?'” Drenon said. “You can go out on the web and find that and compare it to what you want to do.”
Members help make our journalism possible. Become a CSP member today and unlock exclusive benefits, including unlimited access to all of our content. Sign up here.