Due to online competition, retail store margins are already severely strained. Additionally, shrink, or a shortage of inventory relative to the records that are currently available in the system, could negatively impact a retailer's bottom line. Shrinkage is the loss of inventory due to factors including employee theft, shoplifting and other issues. It can happen on the sales floor, at check-out, or even at the store exit.

Since more shrinkage results in lower profitability, shrinkage numbers put merchants at risk. Shrinkage reached an all-time high in 2021, according to the National Retail Security Survey 2021, and it cost retailers 1.6% of their revenue.

Additionally, the drive to digitize stores and provide more frictionless and contactless self-services, such as self-checkout alternatives, is advantageous to both customers and retailers in the post-COVID era. New developments in retail digitization do, however, have the potential to open up new paths for fraud.

An analysis of shrinkage

Reducing shrinkage is the primary objective of loss prevention systems. To keep an eye on goods and employees, many advancements in sales-floor surveillance have been achieved, including cameras, electronic article surveillance (EAS) warnings and radio frequency identification (RFID)-enabled loss prevention. However, they limit retailers' ability to react quickly.

Additionally, there has been an upsurge in employee theft, also referred to as "sweethearting," in which employees give products to friends and family without charging them or bill for cheaper goods than the ones being bought. Depending on the cooperation of an outside shoplifter and a Point of Sale (POS) system employee, many different types of retail theft happen at the register.

Also, the likelihood of theft has increased with self-checkout. In order to prevent scanning the most expensive goods, buyers might place them directly in a bag while scanning other items.

Computer vision and AI can decrease checkout fraud and product loss

Retailers are investing in new technologies to combat retail shrink with stronger loss prevention strategies at the front of the store as a result of these growing issues. In order to help businesses fight shrink, lower theft, and improve inventory management, retail security professionals are utilizing artificial intelligence (AI) and computer vision.

Integration of data from item-level tracking with computer vision and POS can help combat checkout fraud, including at manned registers and self-checkouts. Associates can spot fraud and take immediate action by comparing item-level counts to POS-generated counts. Retailers may improve checkout procedures by making them smarter with the aid of AI and computer vision, which lowers theft and enhances inventory control.

How might this look in practice? In order to detect and stop theft by customers, a retailer could add a camera to the existing checkout lanes and use AI to compare the number of products scanned. The camera deployed sees the items being scanned as a customer scans them at the POS system. The integrated POS system receives the entire item count right away when it is generated. The POS system and the video camera are connected, and after the items are scanned, the POS system adds the camera-generated count to the overall count.

A system can be configured to send a dashboard, POS alert, or mobile alert to the staff member in charge as notification of potential theft or incorrect billing if the two numbers don't match. Before a transaction is processed, it allows store employees to step in and help, enabling the manager to process the transaction manually.

If an AI system flags a suspicious transaction, store staff can inform customers of the system error and give them the option to pay for the unscanned items. Ultimately, the process may help in behavioral change, resulting in reduced shrinkage and cost savings.

Retailers are aggressively investing in technology to combat the surge in organized retail crime, according to the National Retail Security Survey 2021. Product loss and resulting revenue leakage may be impacted by a loss prevention system that uses AI and computer vision.