Shrinkage in retail indicates the loss or reduction in inventory due to theft of items by employee, damage in transit or in store, shoplifting, vendor fraud, and other such mishaps. It is becoming a major concern for retailers as these losses have a negative impact on their bottom line. An estimate from the National Retail Security Survey on retail theft indicates that, in 2016 shrinkage, cost retailers around $49 billion in losses. In 2017, on an average Shrink cost across retailers was about 1.33% of sales. Shrinkage rate in supermarkets tend to be usually higher at around 2.5% of the total sales.

The top 5 types of retail shrinkages include

1. Non-Scanning Loss: This loss occurs when any product goes out without scanning at point of sale (POS) checkout, or the POS operators make mistakes while scanning


2. Basket Based Loss: This loss occurs when customer fails to take out product from trolley to scan at POS. The 3 major types of basket-based losses include bottom of basket, middle of basket and top of basket


3. Sweet-hearting Loss: This type of losses occurs while giving away merchandise in an unauthorized way and without charge, to a “sweetheart” customer (family, fellow employee, friend, etc.)


4. Self-Checkout Loss: this type of losses occur when customers try to bypass the checkout system without scanning the products


5. Poor Demand Planning Loss: This type of losses occurs due to inaccurate replenishment planning, as the products are perishable, and demand is variable

Conclusion

Shrinkage is the most common problem for all the retail stores, which is quite damaging for retailers as it results in loss of inventory as well as money. AI based solutions are being adopted for enhancing the performance of the retailers. AI can help retailers in mitigating shrinkage in following ways:

• Video and data analytics are being used for analysing any fraudulent activities in the retail stores.


• Video intelligence, image processing & recognition algorithms are used for immediately detecting items left anywhere in the cart.


• AI can also analyse the unusual behaviours of its staff, to understand what is going on at the checkout system. It can help in identifying if someone is covering bar code, skipping the scanner and directly taking the merchandise, and more.


• AI can also identify the mismatch between transaction receipt and the item being scanned in the video, by using computer vision technology.

– Sonam Chawla
ICT – Research Analyst
Infoholic Research