Retailers Combine Intelligence with Data Using AI, RFID

Published: September 16, 2024
  • With RFID tags tracking inventory through stores, many retailers are beginning to leverage AI and Generative AI to boost the benefits the technology provides.
  • Initially stores are finding ways to optimize merchandising and improve efficiency by gaining insights about what is selling and what is not, when and where.

For a retailer there is a perfect shopping scenario: a customer walks into a store to browse. They find exactly what they are looking for, in the size that they want, complete their purchase fast, and walk back out of the store. It’s an easy transaction for both seller and buyer, and the satisfied customer can be expected to return.

The reality is that scenarios like this are few and far between.

Retailers have the challenge of understanding when, why and how things become more complicated in their stores, and what they can do to better serve customers at each site. Artificial intelligence (AI) helps, but it is dependent on data from technologies like RFID to build decision making.

Step One: Get Data, Step Two: Understand it

The key benefit RFID has traditionally offered retailers is the capability to track inventory. But the technology use cases are expanding and with AI now providing ways to analyze the data, new applications and benefits are likely to continue, according to Sam Vise, CEO of retail intelligence company Optimum Retailing.

Optimum Retailing recently composed a white paper related to What Retailers Need to Know About AI-powered RFID Right Now. The paper points to the coupling of AI and RFID as a way to revolutionize the entire retail supply chain.

In fact, while RFID tracks unique IDs on products, the read data offers a treasure trove for AI systems that can then begin to extrapolate what the data means, and how it can enable stores to become more profitable, reduce waste and limit labor.

Managing Seasonal, Geographic Differences

By detecting when an RFID tagged product moves from a display for instance, a store’s system can understand which products customers are interacting with, and—based on that information and point of sale data—what is going to sell, where, and at which times, to make sure the right products are on the shelf.

“One of the challenges that retailers face today is how to get the data out of the store in order to make sure that the customer experience is perfect,” said Vise. With AI implemented, “what we can do is a lot of the predictive analysis—what products are selling faster per region, store, season or time of day.”

Retailers with multiple stores know that sales are not often consistent from one site to another. And while they may know that the New York store has sold out of a product faster than its Miami counterpart, they can’t easily know what’s happening in real time, or quickly adjust how they should distribute goods to meet current demand.

By combining RFID data with AI, Vise said, retailers can gain an “understanding of what inventory a store has and how it is being interacted with.” One example would be ensuring that the goods  on display in a store are available but also specific for the audience.

Finding Ways to Localize Offerings

Tech companies are now offering AI as part of an RFID based solution.

Optimum Retailing launched AI as part of its own solution offerings in December 2023, based on the trends observed with customers capturing data from RFID tagged products.

Even before using RFID data, Vise recalled, “we were very effective as a software company to notify stores [and retail management] what they should have on display. We were also very good at capturing the nuanced differences between all the different stores.”

Retailers Combine Intelligence with Data Using AI, RFID is the sixth story in an ongoing series exploring how AI is impacting the RFID and IoT industries.

Bringing AI to Healthcare

AI Drives Efficiency for Automotive Management

Examining The Interdependence of AI and RFID

AI on the Edge

RFID, RTLS Bring Data to AI in the Factory

Now he sees the data from the RFID tags as gifts from heaven, and enablers of AI. The data can be used to collect intelligence around regional and demographic differences between stores, and how products are received in those sites.

“Now that [retailers are] implementing RFID we get visibility of that. We can see what’s on display and then with AI we can start modeling what is happening in terms of how consumers are interacting with their products and how the marketing is working,” Vise said. And if a product isn’t selling the system can further identify whether that product has actually been put on the shelf, indicating it doesn’t impress shoppers, as opposed to being forgotten in a backroom.

Combined Solutions with Image Recognition

Vise pointed to a beauty products chain customer that is implementing AI along with RFID to integrate with their planograms. They plan to use technology such as cameras for image recognition to validate products on the shelf.

“We’re using AI to optimize a display and optimize sales within the store, but with RFID we can start influencing and training our model,” he said, such as identifying how to rearrange the store environment based on how people shop.

Additionally, AI can improve operations based on not just RFID-based inventory data, but results from the point of sales (POS) well as existing planograms intended to increase shopper interest in specific products.

Generative AI for Decision Making

This is a sweet spot for Generative AI, as it can next start making recommendations to retailers. Consider Chat GPT in which a retailer manager can pose questions and gain suggested steps to boost sales.

Furthermore, generative AI can be used for simulating scenarios, said Vahan Simonyan, Optimum Retailing’s chief innovation officer. “It can generate the scenarios to predict [for example] under-stocks and over-stocks” that could occur under certain conditions.

The technology could simulate likely scenarios and then advise on the right decision, he added.

“Help Me Plan Winter Merchandise and Promotions”

Managers can use natural language to ask a variety of questions. Those could include what products would be best for a specific store or season, but also much more detailed queries.

For instance, a user could ask about a display window in their New York City 5th Avenue store, and the number of fixtures that should be on display, as well as how many are actually there.

Optimum Retailing is among the few companies that writes its own AI modeling software and applications specific to retail aimed at integrated RFID and AI for specific applications in the retail environment.

Reducing Waste

RFID use with AI also offers waste reduction, often by large margins. By minimizing overstocks, a retailer can also minimize what gets discarded.

As a retailer, Vise said, “I don’t want to be in a situation where I’m sitting on 50 percent of my inventory at the end of a season, and I either have to go on sale quickly with my winter inventory or I have to move it out of the store because my spring stuff is coming in.”

There is a cost benefit to this as well—less overstock leads to less shipping and the fuel related to that. And retailers are spared from discounting goods when there is a risk of devaluing their brands.

Another challenge for retailers is managing goods that have been purchased online or at the store, and then returned to the store. Such goods often are never resold, because they aren’t returned to the sales floor, Vise explained.

Do it Yourself Options

Retailers already have disparate systems that include RFID and enterprise resource planning (ERP) systems to manage inventory and omnichannel sales, and a variety of merchandising software. Small retailers with one or more stores often need a low-cost solution and they can choose to do it themselves.

One option is solutions such as Amazon Webs Services’ (AWS) recent inclusion of AI for customers, that may not be solely retail focused. The AWS goal is to help customers with little technical experience to use AI to set up their own Internet of Things (IoT) solution.

The company’s IoT Low Code Assistant is a generative AI–powered assistant for software development and can help in modernizing IoT application development improving reliability and security. It understands a user’s code and AWS resources, enabling it to streamline the entire IoT software development lifecycle (SDLC), according to Amazon.

It also offers data analysis and reporting such as Amazon QuickSite.

Whether retailers do it themselves, are seek solution providers—by leveraging RFID tags and AI to manage that raw data—the future of store management and omnichannel sales is likely to benefit.

Learn More:

About the Author: Claire Swedberg