The AI revolution grocery retail customers never asked for (but can’t live without)
Aug 7, 2025 • 9 min
Today, we’re joined by Craig Norman, Product Marketing Manager at RELEX, to explore how AI has quietly transformed grocery retail. Craig will explain how retailers use artificial intelligence to meet growing customer expectations while managing the complex dance of supply chains, demand forecasting, and inventory management. He’ll uncover how AI analyzes thousands of signals to prevent stock-outs and improve operations, all while staying invisible to shoppers who expect their groceries to be available when needed.
Q: When we think about today’s customers, they have a lot less tolerance for mistakes with their groceries. They are frustrated by wrong items being received, missing products, and delayed deliveries. So how does a grocery retailer use AI systems to meet those reliability expectations, and what happens when they don’t?
Craig Norman: That’s a really great question and a great starting point. You’re right. Customers don’t see everything behind the curtain, if you want to think of it that way. So really what retailers have to do is make sure that their AI systems are connected to all of the data sources that they have, so that when they are using them to generate forecasts and plans, they know that the results are reliable and based on as much information as they possibly can be to be as accurate as possible.
Accuracy alone is not enough. You have to do something about it. But that’s the starting point. You need to ensure that you’ve got all that data pulled into the system and connected to make really robust decisions. When you don’t, you open yourself up to gaps, and those gaps show up in empty shelves or missed opportunities.
Q: When we think about how AI can be used, given its capability to process thousands of signals across an operation, can you walk us through some of those signals? We’ve mentioned AI’s ability to process them, but what exactly are these signals, and how do they come together in a real-world scenario?
CN: Right. So, a really obvious example would be promotional planning for a grocery retailer, or another type of retailer, for that matter. When you think of that process today, it can be multiple teams at their own individual parts of the process. How much they need to buy, what’s the price, where’s it positioned in the stores, how does it affect the planogram, all the decisions that go into it. And each of those is a signal.
And so, when you’re using AI in a planning scenario on a system like RELEX, it allows you to operate on a common data source, a single source of truth. By doing that, your supply planners, your category managers, and merchandisers are all using and sharing the same information. So, you can create a much better and more successful promotion that meets your expected demand or can even be tailored down to a store or regional level in a profitable way that grows the business.
Q: In grocery, it’s not hyperbole to say that in grocery timing is everything. Produce spoils, demands shift hourly. There are, especially in the current day, regular supply chain hiccups. So, how has AI changed the speed at which grocery retailers can identify and act on emerging issues and challenges?
CN: So, disruption is the new normal, and that could be tariffs, bird flu, bad crops, or whatever is happening in the world. And these have implications for what you can order and ship into your stores and make available for your shoppers. So, the customer sees that eggs aren’t available or cost much more. They don’t necessarily know why. They only know that they can’t get them for breakfast. So, in this case, with all those disruptions happening, timing is everything.
What AI can do is ingest that sea of data we’ve just talked about and pinpoint where those problems might occur before they happen. It can analyze the trends, look at common themes with specific suppliers or in certain areas, stores, and regions, and help your teams make faster decisions to get past those bottlenecks. It can be as simple as finding an alternative supplier. It could be rerouting trucks to different stores or providing a substitute for a product you can’t get anymore. Regardless of how AI helps you, it can speed up those decisions at a granular level – down to an individual product at a store at a specific time of day, morning or afternoon – and that’s really hard to do manually.
Q: When we look at AI systems in general, and how they fit into a grocery operation, there’s an interesting tension between building these incredibly sophisticated systems and having them in your operation while keeping the operation simple enough for teams to work efficiently. So how do you strike that balance, and what happens when that technology gets too complex for practical use?
CN: I think the really vital thing to say here is that AI is great. It can help you solve many things in ways you couldn’t do manually. But you can’t look at it in a vacuum and say: “We’re going to put it in and hope for the best.” You need to evaluate how you operate your business, your specific requirements, and the processes you have in place. How will the AI impact these? What will change? And you should do a really good job of defining that. And that’s the balance.
Learn More: Scaling AI for consumer-centric growth
A typical person can only deal with so much change at once. So, you want to balance the amount of change you’re pushing through with the concrete capabilities and benefits that the new system will provide. When that gets out of balance, or the technology feels too complex, you risk the problem of your employees not embracing it. And that’s true whether it’s a central planner looking at forecasts or a store employee doing an inventory count. The more complex a new system seems, if it’s adding a step, the less likely teams are to adopt it and continue to use it regularly. And that’s what you don’t want to happen. You need to strike that perfect balance.
Q: Customers really care about getting their groceries quickly and accurately. They don’t care about the AI behind it. So how do you, as a grocery retailer, measure success when your best work is invisible to the end user — the customer?
CN: At a very top level, when you think of the customer experience, there are a couple of ways a retailer could look at this. One is what’s called on-shelf availability – essentially, what’s the percentage chance that an item will be on the shelf when a customer wants it at that location? For online orders that might be picked up in store, retailers may also look at orders fulfilled in full. Did they have the product on the shelf when the shopper came in to buy it or when they came to pick it up? This is a key success metric.
Learn more: Best practices for the retail grocery supply chain
Beyond that, there are traditional business metrics. What is the sales growth that you’re achieving? The profitability? Is it going up or down? How much are you wasting, or how much product is spoiling, and how efficient is your labor in getting work done? Those metrics are crucial whether you’re a grocery retailer or any other retailer. They are the four key levers of business success that, even outside of the discussion of AI, people will be looking at to try to improve their businesses. So, I think you look at those metrics, and then the question becomes: how does the AI system allow you to amplify your goals in ways that manual work can’t achieve?
Q: When we look at the landscape broadly, a fair few major grocery players are investing heavily in AI. Even outside of grocery, it’s a hot technology. People want to implement it and reap the benefits. But with all this implementation and all this interest in AI, how do you separate truly transformative AI from just having AI for the sake of it? What should the industry be focusing on that it’s currently missing?
CN: I think the starting point is actually outside of AI. It’s those big goals that really move the business forward. What pinch points or hurdles would you need to overcome to improve your business, not just by 5% but by 10x, a 100x improvement? Those massively impactful goals and strategies —you start with them. AI needs a goal to achieve anything. AI by itself is just a term. It’s a tool. So, having AI or a chatbot for the sake of having it really going to get you anywhere. How does the AI system enable you to achieve those really big, meaningful goals? So, start with the strategy, then find the tools that let you get there.
I think that one part that gets overlooked is, as people race to implement AI or call things AI, they think they can simply start up a chatbot, which is nice. It helps people ask questions and get an answer more naturally, but they’re not connecting it back to those goals, or they don’t have a big enough goal that will make a sizeable difference in their business.
Then when you look at the AI technology itself, now and for the future, when you’re looking at that strategy, you need to also be able to connect the dots. How do generative AI and agentic AI- these terms that get thrown around a lot now- how do those things connect with the purpose-built machine learning and AI tools that you might have already invested in? How do they work with all the data you already own? And how does that ecosystem of AI, your purpose-built AI with the new generative AI, work together to amplify the results you need to reach your goals?
And your people are still going to be part of this. They’re going to be the ones helping set up those overarching goals and determining whether the goals have been met. So, humans are still there. They’re just allowed to be the strategic partner to technology, enabling them to make faster, more effective decisions.
Q: When we think about the technology that utilizes AI — the platforms that have taken it on. Let’s take RELEX as an example. Could you give us a quick overview of what a platform like RELEX provides a grocery retailer? Suppose a grocery retailer is considering AI-enabled platforms for their operation. What is it that RELEX can actually give them in terms of capabilities?
CN: RELEX has built its capabilities using a few decades of knowledge. I like to say that we know our math. It’s a simple way of putting it. But to really answer that question, RELEX offers a unified platform with comprehensive, market-leading AI and purpose-built solutions that address key problems a grocery retailer would have. How do you plan your demand and forecast more accurately to ensure you have the products you need on the shelf? How does that influence pricing and promotion decisions? How does that all impact the assortment plans you’re working on? Because none of these exist in a vacuum. Somebody’s deciding on what to put on the shelf, the price it needs to be at, when it’s going to be promoted, and your supply chain is executing on that to deliver all of the product to the store. Connecting those activities on a unified platform like RELEX enables faster, more collaborative decision-making with better results. So that’s one area.
The other thing I would say is that RELEX invests a lot of attention in linking emerging technologies and new capabilities to what we already have available. We’re investing in agentic AI, but we’re doing it in a way that connects them into this ecosystem of solutions. So, you’re benefiting from a purpose-built system with automation and the simple interaction of generative AI and agentic AI. You get the best of both worlds. And I think when someone evaluates RELEX, for example, our heritage has great appeal. The connectivity from our unified platform is a differentiator, and the roadmap, the goals, and the vision of where we want the platform to go are what set us apart.
Learn more: AI at RELEX — the right approach
Q: If you had to briefly describe the impact of RELEX, or an AI-enabled platform like RELEX, on the customer experience, how would you succinctly explain that?
CN: I think if we’re talking customer experience, it’s that RELEX helps ensure that products are on the shelf where you need them, when you need them, and in the right quantity to make profitable sales. And that’s it.
Q: What would you want a grocery retailer to this to take away from this discussion?
CN: I think it’s — go into the AI conversation having done some homework on what you really want — your ambitious goals. Then talk to RELEX about those ambitious goals. We have robust solutions across key areas previously mentioned, and delivering measurable value is at our core. It’s our company mantra. Delivering increases in sales, reduction in spoilage, improvement, and profitability. We provide all of this for companies day in and day out. And if I have to say anything to a potential retailer looking at a system like RELEX, it’s: give us a call, because we can deliver those same results for you too.