Pushouts – when thieves posing as shoppers fill carts and baskets with expensive merchandise and push them out the door without paying – cost retailers hundreds of millions of dollars in lost profits every year. According to the National Retail Federation, each pushout averages over $500 in stolen merchandise. Sadly, nearly all of these theft events are undetected, until now.
Artificial intelligence (AI) fundamentally changes how retailers can combat these hard-to-detect theft events. As AI algorithms gain sophistication, and required hardware is ever more affordable, AI is increasingly automating a range of investigative tasks and capturing theft events in real time. By teaching machines to automatically detect theft behaviors, retailers can quickly and cost effectively identify and stop these everyday crimes.
AI is perfectly suited for detecting pushouts. These, as well as other AI-powered tools, are cost effectively changing and increasing the effectiveness of loss prevention tactics. Although each pushout represents a significant loss, the prevention of which is certainly desirable, pushouts occur somewhat infrequently, are often difficult to detect, and even harder to stop. It is simply inefficient and uneconomical to detect and prevent pushouts using humans.
These challenges are readily overcome by AI technology. It is on 24/7, never gets tired, analyzes every event, and can take real time action to stop pushouts. Automating pushout detection empowers retailers to cost effectively fight this crime.
AI delivers several advantages in the pushout battle. First, it identifies the magnitude of the problem. Using AI-based systems, investigators review bookmarked video revealing the frequency of pushouts in stores as well as exactly what merchandise is being stolen. This supports development of increasingly effective product protection strategies.
Second, AI makes real time response a reality. Detected pushouts can instantly alert, making the offender aware that the theft attempt has been detected and the event captured on video. An audible message and visual alarm often result in the offender abandoning the theft. Since pushouts are often committed by repeat offenders, who have grown comfortable stealing from the same store time after time, this deterrence action also stops multiple future thefts.
A working example of AI fighting pushout theft is incorporated into SmartDepart™, a new system from Indyme Solutions. In this example, harnessing the power of AI turns ordinary LP cameras into powerful pushout-prevention tools. SmartDepart detects shopping carts and baskets approaching the exit and uniquely identifies those that bypassed checkout. AI algorithms also detect merchandise in baskets and carts and track their travel path as they exit the sales floor and approach the exit.
A merchandise-laden cart or basket detected as bypassing checkout defines a pushout attempt, triggering Indyme’s real-time Active Deterrence actions. Typically, a lightbar above the exit flashes and a customizable voice down-messages the offender. Store associates or LP professionals can also be notified and video surveillance systems bookmarked.
“SmartDepart represents a new generation of AI-based loss prevention tools,” said Joe Budano, Indyme’s CEO, “using ordinary LP cameras, we dramatically reduce the cost of effective pushout protection, especially compared to more traditional locking-wheel solutions.”
Significantly lower deployment costs mean more retailers can afford to protect more stores against insidious pushout thefts. It’s just one example of how a well-executed solution leveraging the power of AI broadens a retailer’s options for solving in-store problems to improve the customer experience, increase staff productivity, and drive down theft.
