Retail’s Big Show is back in force: 3 takeaways from NRF 2023
After a three-year hiatus, NRF 2023: Retail’s Big Show was back in full force this January featuring the best of retail and retail tech. 35,000 attendees. 175 sessions. 1,000+ exhibitors. The buzz was intoxicating as positive energy from retailers and technology partners filled the air in spite of recent economic headwinds. Amazon Web Services (AWS) was excited to participate in the return of the in-person show.
At this year’s Big Show, we saw a huge spike in content that targeted and featured fast moving consumer goods (FMCG) retailers. The floor was booming with technology highlighting the best of store and digital solutions powering retail of all shapes and sizes.
Visiting the Javits Center’s exhibit floors, the AWS Partner team met dozens of industry partners and attended many sessions. Each experience introduced us to new and emerging technologies covering the entire retail solution stack. From our conversations with many AWS customers and partners at the conference, we saw three recurring themes that we believe will have game-changing impact on the show, attendees, and ultimately the future of retail.
#1 – Experience and engagement everywhere
It’s very clear that as an industry, we are way beyond talking about traditional point-of-sale (POS) as a differentiating technology. POS is now table stakes. Retailers need – and shoppers demand – entirely new solutions for driving a journey-oriented and experience-focused customer engagement strategy.
What are you doing to make your customer touchpoints more compelling? Are you creating an experience that makes your customers want to come back again and again? These are questions the NRF retail event attendees are trying to address. We’ve been excited to see how AWS Partners like Mad Mobile are helping major retailers and restaurants modernize POS flow, improve customer experience, and optimize labor efficiency. Yet how will retailers harness continuously changing technology in a way that it can be delivered quickly – and defect-free?
The answer is a composable approach. We saw more AWS Partners and fellow MACH Alliance members (including Fabric, Constructor.io, VTEX, Fluent Commerce, Pivotree, and commercetools, and more) talk about adopting composable commerce and saying goodbye to monolithic architectures. The microservices architecture approach has spread beyond ecommerce and is now changing enterprise software development and the entire IT landscape. At NRF, we were excited to visit the MACH Haus and lead a standing-room-only session with the Fabric and Chico’s FAS teams. Adopting MACH allows you to move away from a cycle of constant re-platforming and leverage a best-of-breed strategy that delivers better speed to market, reduced risk, and a path to seamless customization that allows you to innovate without the hassle of disruptive upgrades.
Finally, we can’t talk about experience and engagement without talking about frictionless shopping. This year’s show featured over 30 vendors offering frictionless payments solutions – and that’s just payments! Everywhere you turn these days, there is another company claiming to have invented the next big thing in checkout since the UPC barcode. Thousands of conference-goers stopped by the AWS booth to see our Just Walk Out by Amazon, Dash Cart, and Amazon One technologies and experience a retail experience that is easy, differentiated, and frictionless. Just Walk Out by Amazon is proven technology that is used in Amazon Go, Amazon Fresh, and Whole Foods Market locations as well as other retailers in the US and Europe. The diagram below illustrates the simplicity of the Just Walk Out customer experience.
#2 – Cost reduction, cost reduction, cost reduction
While the mood at NRF was upbeat, the somewhat shaky economic environment highlighted the necessity of cost reduction and was top of mind for many attendees. The big question for retailers: How will they bring these solutions into their stores? And, how they will they optimize their workforce to control costs while also delivering a top-drawer consumer experience? Do they need to be fully autonomous? Are they integrating all their platforms to make it easier for employees to do their jobs?
Cost reduction used to require looking under every rock to find savings. Better labor and inventory management added to the savings, but those only addressed about 80 percent of it. So, what about the other 20 percent? It’s not as easy anymore. But there are at least three areas where you can find those extra savings: supply chain optimization, advanced price optimization, and options for addressing repetitive tasks and improving labor management in the store.
From a supply standpoint, advanced analytics and better metrics are keys to improving visibility into your inventory and supply chain and helping you optimize your warehousing and logistics operations. Industry leaders, such as Fluent Commerce, have worked with AWS to provide a best-of-breed order management system that helps retailers of all sizes squeeze savings from their supply chain while also meeting their customer’s growing appetite for getting the products they want right now at the best possible price.
As retailers strive for the elusive infinite shelf, they must keep a constant eye on costs that creep in from every direction. While FMCGs are sold at a relatively low cost, many also carry a relatively low margin. So, even small increases to drayage (short-haul transport) can quickly swing gross profit below your target levels. This is where price optimization enters the fray. Innovators, such as Peak.ai, use advanced decision intelligence that combines data and artificial intelligence (AI) to help a leading UK multichannel retailer retain $3 million in additional margin through markdown optimization.
So, while optimizing supply chains and price optimization are solutions above the store, there is a whole bevy of mundane tasks on the ground that stores have to deal with every day. Everything from stocking, to cleaning, bagging, and let’s not forget, scanning product at the till. How do you address all of them cost-effectively? While we haven’t exactly seen robots stocking shelves in a grocery store, that’s closer to reality than you might think. Take, for example, Marty, a clean-up assisting robot that was introduced into Giant Food and Stop and Shop locations back in 2019. Marty glides through the aisles looking for clean-up opportunities. Kids love Marty. But how long before Marty is asked to inform computer-aided-ordering (CAO) systems about stockouts (if that’s not already happening)? Will Marty grow arms and begin filling those empty shelves? The future is always in motion and Marty is an example of what’s to come. We may all want to get ready for more labor-saving robots coming soon to a supermarket near you.
At this year’s show, we saw countless solutions for improving in-store operations. For example, what can you use to optimize queues at checkout? What about stockouts (assuming you don’t use robots to roam the aisles)? How about recognizing products that are in the bottom basket of a shopping cart? Having eyes on shelves, queues, and baskets is nearly impossible for store associates who are already busy stocking, cleaning, and helping or checking out customers.
One of the answers is to employ computer vision (CV) in your store. Computer vision at the edge is here. AWS Panorama, for example, allows you to add computer vision to your existing fleet of cameras or augment your current in-store cameras with AWS Panorama devices. The solution “takes the blinders off” and helps you see more of what’s happening in your store, even while you’re asleep. It also integrates seamlessly with your store’s local area network. Here is a typical deployment model, below.
With CV, you can target your highest value-capture opportunities by employing AI at the edge. To do this, you can build your own custom CV models using Amazon SageMaker or perhaps get started with models developed by AWS Partners that are working with AWS Panorama today. Those models can be deployed on-premises to an AWS Panorama appliance that can use existing IP cameras as well as AWS Panorama-enabled devices. Data derived from these CV models can give you insights into trends that may be specific to only a few locations, but can directly influence how they perform.
#3 – Machine learning is going mainstream
We have watched machine learning (ML) evolve over the last decade and it has now become part of our everyday language. While no longer the stuff of science fiction, ML isn’t quite in our rear-view mirror either. Granted, when we are faced with an avalanche of data, the first thing we ask today is if there’s something valuable buried inside that an ML algorithm could unearth. But the truth is that we – or rather, our machines – are still learning. Each iteration of a ML routine reveals new patterns that we humans could never imagine.
How can ML help us? First, it can help us distinguish between FMCGs and everything else. Because of their high consumer demand (think soft drinks, milk, eggs, beer, and over-the-counter drugs like aspirin), FMCGs have a short shelf life, accounting for over half of consumer spending. Hence, all eyes – and ML solutions – are focused on FMCG. Not surprisingly, FMCG sales patterns are well-studied and causal factors tend to be better understood and predictable. ML is remarkably good at predicting future sales based on national advertising and other highly directive influences. ML can combine external factors to create uncannily accurate forecasts of demand for FMCG, and many retailers are employing this practice today.
But what about slow-moving consumer goods? These are the ones that sit on the shelf long enough to collect a bit of dust. But then, something happens. Out-of-the-blue, a dusty item might sell out in one location. Take capers, for example. These tiny, pickled fruits are used in a handful of dishes year-round and they’re found close to olives in most grocery stores. Capers typically don’t move very fast. However, their movement can be influenced by micro-causal factors like an episode on a popular cooking channel that features chicken piccata, or a location-based event like an in-store demonstration that shows how capers can be used to substitute for anchovies in a salad dressing. Suddenly, very quietly, every last bottle of capers flies off the shelf – but only in a few locations, while the rest of the chain has plenty on-hand. What do you do about that? It’s only capers, so why does it matter? It matters because capers are one of hundreds of products influenced in the same way – and if you aren’t paying attention, you’re losing sales. And who can pay attention to every item in every location of your chain? (Hint: ML can.)
Sales patterns tied to micro-causal factors – including specific locations, the weather, cultural events, and more – are hard to recognize without the help of ML. And even then, it can be hit-and-miss. But with AWS Solutions, such as Improving Forecast Accuracy with Machine Learning, the walls are coming down. If you think of every item in every location having multiple factors influencing future sales – and a typical grocery store’s assortment tops 80,000 active items – the scale of the opportunity becomes evident. No human can see the patterns and we need sophisticated ML algorithms to do it for us.
AWS offers solutions for helping you with forecasting, such as Amazon Forecast, and AWS Partners, like Anaplan and its PlanIQ solution that can help improve assortment planning by leveraging ML with internal, external, and third-party data to unlock insights. Prepared with these tools, you will gain the product visibility necessary to optimize replenishment levels and react quickly to changes in demand. As the evolution continues in ML, it will be amazing to see the options and innovations in this space for 2023 and beyond.
Bringing it together
Born from retail, built for retailers, AWS is continuously developing industry-specific services that are market-tested by Amazon.com and then offered for customer use. AWS uniquely combines the agility of a startup with the experience of an enterprise-class retail leader. This experience enables us to deliver massive growth for some of the largest retailers in the world while leveraging the expertise and resources of 100,000 companies from more than 150 countries in the AWS Partner Network.
It was a joy seeing everyone we connected with in New York City. We look forward to taking this positive momentum into 2023 to help build a more engaging, experiential, operationally efficient, and innovative industry.
Learn more about how AWS and our partners can empower you to create exceptional experiences built for the future of retail.