The AI Revolution in Grocery

Taking a Foundational Look 

AI is easy enough to understand on a surface level: It’s the capability of computer systems or algorithms to imitate intelligent human behavior. It uses massive amounts of data to make decisions and solve problems on behalf of its user, and it can make forecasting predictions in business use cases.

According to Lori Schafer, CEO of Ponte Vedra Beach, Fla.-based Digital Wave Technology, the whole idea of AI is to automate processes and make things much faster and more efficient than a human could. “Data is the fuel, if you will, for any type of AI solution, and it’s the job of the AI solution to be able to take in that clean data, and then look at every possible way to automate that data and to also personalize the data as those AI solutions are used to benefit the business,” she explains.

Schafer stresses the importance of delineating between standard AI and its new iteration of GenAI when discussing the finer points of each. Numeric or analytical AI works on numbers, while GenAI works on text, video and voice. Either way you slice it, AI in food retail holds the promise of streamlining mundane tasks, cutting food waste, boosting the customer experience and reducing operational costs.

In today’s grocery ecosystem, Schafer says that AI is most commonly used for inventory management and optimization, which allows grocers to more efficiently predict and forecast their demand, and then make sure that the inventory they keep in stock is optimized to that demand curve. As for GenAI, current common uses include chatbots that answer shopper questions online, email and promotion personalization at mass scale, and content management that allows brands to write product descriptions in their own voice, for example.

Paul Pallath, VP of applied AI at Houston-based consultancy Searce, shares that supply chain optimization is another business function getting a boost from GenAI. As grocers work with multiple vendors, generate copious amounts of documentation and work to reconcile invoices against purchase orders, GenAI can help take the guesswork out of parsing that documentation. 

Other common uses explored in a recent Coresight Research report include computer vision and AI-powered, fully autonomous checkout systems and smart carts; warehouse automation; and video commerce. Sprouts Farmers Market, for example, has partnered with RELEX Solutions to optimize its forecasting and replenishment processes to minimize spoilage and improve on-shelf availability. 

[RELATED: RELEX's Partnership Optimizes End-to-End Retail Ops and Inventory Planning]

The Kroger Co., meanwhile, employs technology from Ocado that uses AI algorithms and machine learning to optimize the picking and packing operations of online orders. That automation reduces labor costs and improves speed and accuracy.

Taking the Leap

Despite the promise of AI in food retail, there remains a gap between those grocers harnessing its power and those still waiting to make a move on the technology. While this is often directly related to the size of a company and its available resources, Schafer recommends being very sharp on the value proposition of AI when looking to integrate it into everyday operations.

“I think it needs to be simplified, and what I would suggest to any independent retailer, or even a regional chain for that matter, is [to have a conversation] at the CEO level, ideally a CIO, CEO or CFO, because they know where the needs are,” she advises. 

Starting small and working toward identifiable pain points or goals can go a long way. Setting up AI for content management, such as writing product content or recipes, for example, is an easy place to start that offers tangible value to the organization and minimizes the difficulty of change management.

Regardless of speed to integration, a good place to start, and often the real key to AI success, is the use of clean data, according to Schafer. 

“One of the biggest things we say when we get started with retailers in terms of adoption is ‘let’s get the data out of the silos,’ and we can put it into a common data lake where all the important data is gathered together,” she explains. “Then we can automatically cleanse that data because sometimes you have a lot of misspellings or duplicates. Whenever human beings enter data, there’s going to be mistakes, so the first step is to integrate the data.”

Pallath is on the same page in terms of employing the best data possible, stressing that AI is only as good as the data that you collect. “Grocers are not technology companies, so their maturity curve in terms of having the data ready for AI is something that we see they struggle with,” he says. 

[RELATED: SPECIAL REPORT: Retail Innovation Outlook]

Continues Pallath: “While they have ideas in terms of what they want out of AI, they’re not ready to get there. Much of our work starts with helping them with the infrastructure or helping with the data modernization techniques before we land with AI and skills.” 

This is often where having a technology partner can make a huge difference in the uptake. “Partner with a consulting organization who has worked with large grocers to understand the space really well,” advises Pallath.

The Future is Bright

At the end of the day, grocery store operations ranging from seamless checkout to product optimization planning, warehouse operations, and even creative processes like product description and digital advertisement generation are all getting more efficient, thanks to AI. According to Schafer, the technology can easily create time efficiencies of up to 95%.

Pallath believes that the next five years will bring an abundance of advancements in grocery-related AI that will help remove friction from operations, including the ability for individual stores to get a better view into what’s happening on the floor, as well as what’s on the shelf. AI and GenAI will also begin working more in tandem to create further optimizations in terms of inventory, shelf stock, planograms and more, which will translate to a better customer experience and more profit for the retailer.

Schafer thinks AI will become much more embedded in daily operations over the next five years, with automation and personalization being important pieces of the puzzle. Regardless, she stresses that these applications will not take away the need for a human touch within the AI loop.

“Artificial intelligence is not intelligent,” she says. “It’s a bunch of mathematical algorithms that are used to come up with an optimal answer or to process data faster. But the human is always there and needs to be there to shape the message of what AI needs to do, and the human should always approve it. We’re a long way off from the human being out of the loop.”