On a busy Saturday, a shopper at a clothing store finds the right jacket but not the right size. The assistant checks the store system. It says one method is available. But the jacket is not on the rack. It may have been in the fitting room, folded on the wrong table, sitting in the inventory room, or disappeared due to theft that was never reflected in the inventory file. For the shopper, they are out of stock. For the store manager, it’s a lost sale. For HQ, it probably still looks like stock.
This is the gap Spencer HewittFounder and CEO of radarspent years trying to close. “I chose to focus on inventory because this is a product where you can generate a significant amount of value for retailers out of the gate,” Hewitt told PYMNTS CEO Karen Webster, explaining why RADAR moved from its original idea of autonomous checkout toward inventory information.
The timing is extraordinarily appropriate. RADAR recently raised $170 million in Series B funding, co-led by Gideon Strategy Partners and Nimble Partners with participation from Align Ventures, bringing the company’s valuation to $1 billion.
The company’s pitch is straightforward: Physical stores have long lacked the data layer that e-commerce takes for granted. RADAR says it provides real-time accuracy for 80% of the commerce that still occurs in stores, with 99% item-level inventory accuracy and continuous visibility into where products are located and whether they are in stock. Deployed in more than 1,400 stores, including American Eagle and Old Navy, RADAR is trying to turn the store into a live checkout system at the same moment when retailers are under pressure to operate leaner, fulfill online orders from stores and use artificial intelligence in practical ways.
The simple truth about data
The business model is based on a simple premise: AI is only as good as the data it can use. Retailers have no shortage of AI pilots, but many still rely on manual inventory, outdated inventory files, and warehousing systems that can’t pinpoint exactly where an item is located. Vertical integration of RADAR is important here. The company provides proprietary overhead sensors, software and analytics as a single system rather than as a loose collection of tools. Ceiling-mounted sensors read tagged items across the sales floor, stockroom and fitting rooms, taking a complete snapshot of inventory every eight seconds and translating raw signals into tasks: restock this item, route this order from this store, flag this lost product, fix this misplaced display.
Advertisement: Scroll to continue
RADAR+ expands this foundation of “what do we have?” to “What’s going on?” The AI analytics platform is designed to show retailers behavior at the SKU level: which items are being picked up by customers, which are going into fitting rooms, which are turning into sales, and which are staying in the wrong place for too long. It also recommends replenishment priorities, alerts employees when product is out of place and gives company teams frequent updates on dwell time, availability on the floor, and recovery time. In other words, it gives the store some of the feedback loop that websites always provide.
The near-term question is what retailers value more: better inventory management or loss prevention. Hewitt’s answer is that the two are closely related. Theft is important not only because the product has disappeared, but because the system may still believe it is available.
“So you could be out of inventory for a whole year,” he told Webster, describing how stolen merchandise can remain as dummy inventory until the next actual count is done. But when retailers buy Radar, the most obvious return is usually a reduction in inventory. “Once you anticipate and prevent stock-outs, you sell more product and you sell it at a higher margin,” he said. The first use case is availability; Subsequent uses include assortment, business, fulfillment, and marketing.
Ask better questions
This makes RADAR an AI story, but not in the abstract sense. Hewitt described RADAR+ as a way to display information that might require people to monitor the store “24/7.” This is easy to do online because the company can create software logs, he said. “We’re doing that now in the store.” Once these records exist, retailers can ask better questions. How many times has this jacket been picked up? What did you walk into the fitting room for? What did he win or lose? What product is strong but sits in a weak spot? The next step is prescriptive: not just showing what happened, but recommending where to place items to increase conversion.
Hewitt’s longer view is more ambitious. Within three years, he expects stores to become more contextual. Recommendations can change based on what is already in stock. Sales associates can have tools to help them advise shoppers in the aisle. Returns could become more automated. Exit can start to fade into the background.
He also sees a future in which the wall between shopping and the physical world becomes thinner: If augmented reality takes hold, the world itself may become more shoppable.
Webster ended the conversation by bringing up the counterpoint that every retailer should consider. A lot of shopping revolves around physical Serendipity. The AI is often instructed: Find me a blue jacket, and he finds a blue jacket. But stores also help people discover what they didn’t know they wanted until they saw it. She suggested that radar could create a “more efficient layer of serendipity.”
Hewitt’s answer was simple: “Absolutely.” He pointed to the renewed interest in businesses in the physical world and the fact that younger consumers are still showing up in stores and in movies. The goal isn’t to make stores look like websites. It’s giving retailers the precision of online commerce without taking away the surprise, touch and discovery that make physical retail worth the trip.
For more video interview:
- Hewitt’s background: He discusses selling designer bags on eBay as a teenager, interning at eBay and how early exposure to e-commerce shaped his question of why brick-and-mortar stores haven’t progressed as quickly as online stores.
- Jay Schottenstein as an Investor: Webster asks about Schottenstein as an experienced retail trader and investor. Schottenstein pushed RADAR to continue thinking about consumer-facing uses and the value of a single platform across the store, Hewitt says.
- Why he started RADAR: Hewitt explains that he originally wanted to solve the problem of checkout lines, but pivoted after learning how much value retailers could gain from optimizing inventory pricing and realizing that inventory information could create immediate value.
