> That analysis showed that self-checkout led to a shrink rate more than 16 times higher than traditional cashier lines. Nearly 7% of self-checkout transactions had at least some amount of partial shrink, compared to 0.32% with cashiers. The analysis also suggested a shrink rate of 3.5% for self-checkout machines, while conventional cashiers only saw a 0.21% shrink rate.
> For its study, the company used computer vision to analyze nearly 5,000 retail transactions, comparing items the shoppers picked up during their shopping trip with transaction data to see what was actually purchased
Then
> According to Grabango, checkout-free technology powered by computer vision can help eliminate self-checkout shrink by tracking what shoppers pick up and charging them exactly what they owe. The company also stressed that average supermarkets could increase bottom-line profits by more than 50% per year by eliminating partial shrink from self-checkout alone.
First, I don’t trust a company selling something to say they did a study, literally using that product, and the result of the study was to sell you something new. Not only that, but everyone knows computer vision isn’t always perfect, it’s similar to LLM AI in that when it works it’s amazing but it’s hard to get working. For this study it seems like they trusted their CV product to be perfect and compared it to the human scanned item list and assumed that was inaccurate.
This article is literally just an ad, not a reliable source!
